src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_double/nb_kernel_ElecCSTab_VdwNone_GeomW3W3_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 #ifdef HAVE_CONFIG_H
  39 #include <config.h>
  40 #endif
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "types/simple.h"
  46 #include "gromacs/math/vec.h"
  47 #include "nrnb.h"
  48
  49 #include "gromacs/simd/math_x86_avx_128_fma_double.h"
  50 #include "kernelutil_x86_avx_128_fma_double.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwNone_GeomW3W3_VF_avx_128_fma_double
  54  * Electrostatics interaction: CubicSplineTable
  55  * VdW interaction:            None
  56  * Geometry:                   Water3-Water3
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecCSTab_VdwNone_GeomW3W3_VF_avx_128_fma_double
  61                     (t_nblist                    * gmx_restrict       nlist,
  62                      rvec                        * gmx_restrict          xx,
  63                      rvec                        * gmx_restrict          ff,
  64                      t_forcerec                  * gmx_restrict          fr,
  65                      t_mdatoms                   * gmx_restrict     mdatoms,
  66                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  67                      t_nrnb                      * gmx_restrict        nrnb)
  68 {
  69     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  70      * just 0 for non-waters.
  71      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
  72      * jnr indices corresponding to data put in the four positions in the SIMD register.
  73      */
  74     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  75     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  76     int              jnrA,jnrB;
  77     int              j_coord_offsetA,j_coord_offsetB;
  78     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  79     real             rcutoff_scalar;
  80     real             *shiftvec,*fshift,*x,*f;
  81     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  82     int              vdwioffset0;
  83     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  84     int              vdwioffset1;
  85     __m128d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  86     int              vdwioffset2;
  87     __m128d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  88     int              vdwjidx0A,vdwjidx0B;
  89     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  90     int              vdwjidx1A,vdwjidx1B;
  91     __m128d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
  92     int              vdwjidx2A,vdwjidx2B;
  93     __m128d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
  94     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  95     __m128d          dx01,dy01,dz01,rsq01,rinv01,rinvsq01,r01,qq01,c6_01,c12_01;
  96     __m128d          dx02,dy02,dz02,rsq02,rinv02,rinvsq02,r02,qq02,c6_02,c12_02;
  97     __m128d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  98     __m128d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
  99     __m128d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
 100     __m128d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 101     __m128d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
 102     __m128d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
 103     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
 104     real             *charge;
 105     __m128i          vfitab;
 106     __m128i          ifour       = _mm_set1_epi32(4);
 107     __m128d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF,twovfeps;
 108     real             *vftab;
 109     __m128d          dummy_mask,cutoff_mask;
 110     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
 111     __m128d          one     = _mm_set1_pd(1.0);
 112     __m128d          two     = _mm_set1_pd(2.0);
 113     x                = xx[0];
 114     f                = ff[0];
 115
 116     nri              = nlist->nri;
 117     iinr             = nlist->iinr;
 118     jindex           = nlist->jindex;
 119     jjnr             = nlist->jjnr;
 120     shiftidx         = nlist->shift;
 121     gid              = nlist->gid;
 122     shiftvec         = fr->shift_vec[0];
 123     fshift           = fr->fshift[0];
 124     facel            = _mm_set1_pd(fr->epsfac);
 125     charge           = mdatoms->chargeA;
 126
 127     vftab            = kernel_data->table_elec->data;
 128     vftabscale       = _mm_set1_pd(kernel_data->table_elec->scale);
 129
 130     /* Setup water-specific parameters */
 131     inr              = nlist->iinr[0];
 132     iq0              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+0]));
 133     iq1              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
 134     iq2              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
 135
 136     jq0              = _mm_set1_pd(charge[inr+0]);
 137     jq1              = _mm_set1_pd(charge[inr+1]);
 138     jq2              = _mm_set1_pd(charge[inr+2]);
 139     qq00             = _mm_mul_pd(iq0,jq0);
 140     qq01             = _mm_mul_pd(iq0,jq1);
 141     qq02             = _mm_mul_pd(iq0,jq2);
 142     qq10             = _mm_mul_pd(iq1,jq0);
 143     qq11             = _mm_mul_pd(iq1,jq1);
 144     qq12             = _mm_mul_pd(iq1,jq2);
 145     qq20             = _mm_mul_pd(iq2,jq0);
 146     qq21             = _mm_mul_pd(iq2,jq1);
 147     qq22             = _mm_mul_pd(iq2,jq2);
 148
 149     /* Avoid stupid compiler warnings */
 150     jnrA = jnrB = 0;
 151     j_coord_offsetA = 0;
 152     j_coord_offsetB = 0;
 153
 154     outeriter        = 0;
 155     inneriter        = 0;
 156
 157     /* Start outer loop over neighborlists */
 158     for(iidx=0; iidx<nri; iidx++)
 159     {
 160         /* Load shift vector for this list */
 161         i_shift_offset   = DIM*shiftidx[iidx];
 162
 163         /* Load limits for loop over neighbors */
 164         j_index_start    = jindex[iidx];
 165         j_index_end      = jindex[iidx+1];
 166
 167         /* Get outer coordinate index */
 168         inr              = iinr[iidx];
 169         i_coord_offset   = DIM*inr;
 170
 171         /* Load i particle coords and add shift vector */
 172         gmx_mm_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 173                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 174
 175         fix0             = _mm_setzero_pd();
 176         fiy0             = _mm_setzero_pd();
 177         fiz0             = _mm_setzero_pd();
 178         fix1             = _mm_setzero_pd();
 179         fiy1             = _mm_setzero_pd();
 180         fiz1             = _mm_setzero_pd();
 181         fix2             = _mm_setzero_pd();
 182         fiy2             = _mm_setzero_pd();
 183         fiz2             = _mm_setzero_pd();
 184
 185         /* Reset potential sums */
 186         velecsum         = _mm_setzero_pd();
 187
 188         /* Start inner kernel loop */
 189         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
 190         {
 191
 192             /* Get j neighbor index, and coordinate index */
 193             jnrA             = jjnr[jidx];
 194             jnrB             = jjnr[jidx+1];
 195             j_coord_offsetA  = DIM*jnrA;
 196             j_coord_offsetB  = DIM*jnrB;
 197
 198             /* load j atom coordinates */
 199             gmx_mm_load_3rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 200                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 201
 202             /* Calculate displacement vector */
 203             dx00             = _mm_sub_pd(ix0,jx0);
 204             dy00             = _mm_sub_pd(iy0,jy0);
 205             dz00             = _mm_sub_pd(iz0,jz0);
 206             dx01             = _mm_sub_pd(ix0,jx1);
 207             dy01             = _mm_sub_pd(iy0,jy1);
 208             dz01             = _mm_sub_pd(iz0,jz1);
 209             dx02             = _mm_sub_pd(ix0,jx2);
 210             dy02             = _mm_sub_pd(iy0,jy2);
 211             dz02             = _mm_sub_pd(iz0,jz2);
 212             dx10             = _mm_sub_pd(ix1,jx0);
 213             dy10             = _mm_sub_pd(iy1,jy0);
 214             dz10             = _mm_sub_pd(iz1,jz0);
 215             dx11             = _mm_sub_pd(ix1,jx1);
 216             dy11             = _mm_sub_pd(iy1,jy1);
 217             dz11             = _mm_sub_pd(iz1,jz1);
 218             dx12             = _mm_sub_pd(ix1,jx2);
 219             dy12             = _mm_sub_pd(iy1,jy2);
 220             dz12             = _mm_sub_pd(iz1,jz2);
 221             dx20             = _mm_sub_pd(ix2,jx0);
 222             dy20             = _mm_sub_pd(iy2,jy0);
 223             dz20             = _mm_sub_pd(iz2,jz0);
 224             dx21             = _mm_sub_pd(ix2,jx1);
 225             dy21             = _mm_sub_pd(iy2,jy1);
 226             dz21             = _mm_sub_pd(iz2,jz1);
 227             dx22             = _mm_sub_pd(ix2,jx2);
 228             dy22             = _mm_sub_pd(iy2,jy2);
 229             dz22             = _mm_sub_pd(iz2,jz2);
 230
 231             /* Calculate squared distance and things based on it */
 232             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 233             rsq01            = gmx_mm_calc_rsq_pd(dx01,dy01,dz01);
 234             rsq02            = gmx_mm_calc_rsq_pd(dx02,dy02,dz02);
 235             rsq10            = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
 236             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
 237             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
 238             rsq20            = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
 239             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
 240             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
 241
 242             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 243             rinv01           = gmx_mm_invsqrt_pd(rsq01);
 244             rinv02           = gmx_mm_invsqrt_pd(rsq02);
 245             rinv10           = gmx_mm_invsqrt_pd(rsq10);
 246             rinv11           = gmx_mm_invsqrt_pd(rsq11);
 247             rinv12           = gmx_mm_invsqrt_pd(rsq12);
 248             rinv20           = gmx_mm_invsqrt_pd(rsq20);
 249             rinv21           = gmx_mm_invsqrt_pd(rsq21);
 250             rinv22           = gmx_mm_invsqrt_pd(rsq22);
 251
 252             fjx0             = _mm_setzero_pd();
 253             fjy0             = _mm_setzero_pd();
 254             fjz0             = _mm_setzero_pd();
 255             fjx1             = _mm_setzero_pd();
 256             fjy1             = _mm_setzero_pd();
 257             fjz1             = _mm_setzero_pd();
 258             fjx2             = _mm_setzero_pd();
 259             fjy2             = _mm_setzero_pd();
 260             fjz2             = _mm_setzero_pd();
 261
 262             /**************************
 263              * CALCULATE INTERACTIONS *
 264              **************************/
 265
 266             r00              = _mm_mul_pd(rsq00,rinv00);
 267
 268             /* Calculate table index by multiplying r with table scale and truncate to integer */
 269             rt               = _mm_mul_pd(r00,vftabscale);
 270             vfitab           = _mm_cvttpd_epi32(rt);
 271 #ifdef __XOP__
 272             vfeps            = _mm_frcz_pd(rt);
 273 #else
 274             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 275 #endif
 276             twovfeps         = _mm_add_pd(vfeps,vfeps);
 277             vfitab           = _mm_slli_epi32(vfitab,2);
 278
 279             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 280             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 281             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 282             GMX_MM_TRANSPOSE2_PD(Y,F);
 283             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 284             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 285             GMX_MM_TRANSPOSE2_PD(G,H);
 286             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 287             VV               = _mm_macc_pd(vfeps,Fp,Y);
 288             velec            = _mm_mul_pd(qq00,VV);
 289             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 290             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq00,FF),_mm_mul_pd(vftabscale,rinv00)));
 291
 292             /* Update potential sum for this i atom from the interaction with this j atom. */
 293             velecsum         = _mm_add_pd(velecsum,velec);
 294
 295             fscal            = felec;
 296
 297             /* Update vectorial force */
 298             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 299             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 300             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 301
 302             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
 303             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
 304             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
 305
 306             /**************************
 307              * CALCULATE INTERACTIONS *
 308              **************************/
 309
 310             r01              = _mm_mul_pd(rsq01,rinv01);
 311
 312             /* Calculate table index by multiplying r with table scale and truncate to integer */
 313             rt               = _mm_mul_pd(r01,vftabscale);
 314             vfitab           = _mm_cvttpd_epi32(rt);
 315 #ifdef __XOP__
 316             vfeps            = _mm_frcz_pd(rt);
 317 #else
 318             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 319 #endif
 320             twovfeps         = _mm_add_pd(vfeps,vfeps);
 321             vfitab           = _mm_slli_epi32(vfitab,2);
 322
 323             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 324             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 325             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 326             GMX_MM_TRANSPOSE2_PD(Y,F);
 327             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 328             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 329             GMX_MM_TRANSPOSE2_PD(G,H);
 330             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 331             VV               = _mm_macc_pd(vfeps,Fp,Y);
 332             velec            = _mm_mul_pd(qq01,VV);
 333             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 334             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq01,FF),_mm_mul_pd(vftabscale,rinv01)));
 335
 336             /* Update potential sum for this i atom from the interaction with this j atom. */
 337             velecsum         = _mm_add_pd(velecsum,velec);
 338
 339             fscal            = felec;
 340
 341             /* Update vectorial force */
 342             fix0             = _mm_macc_pd(dx01,fscal,fix0);
 343             fiy0             = _mm_macc_pd(dy01,fscal,fiy0);
 344             fiz0             = _mm_macc_pd(dz01,fscal,fiz0);
 345
 346             fjx1             = _mm_macc_pd(dx01,fscal,fjx1);
 347             fjy1             = _mm_macc_pd(dy01,fscal,fjy1);
 348             fjz1             = _mm_macc_pd(dz01,fscal,fjz1);
 349
 350             /**************************
 351              * CALCULATE INTERACTIONS *
 352              **************************/
 353
 354             r02              = _mm_mul_pd(rsq02,rinv02);
 355
 356             /* Calculate table index by multiplying r with table scale and truncate to integer */
 357             rt               = _mm_mul_pd(r02,vftabscale);
 358             vfitab           = _mm_cvttpd_epi32(rt);
 359 #ifdef __XOP__
 360             vfeps            = _mm_frcz_pd(rt);
 361 #else
 362             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 363 #endif
 364             twovfeps         = _mm_add_pd(vfeps,vfeps);
 365             vfitab           = _mm_slli_epi32(vfitab,2);
 366
 367             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 368             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 369             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 370             GMX_MM_TRANSPOSE2_PD(Y,F);
 371             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 372             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 373             GMX_MM_TRANSPOSE2_PD(G,H);
 374             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 375             VV               = _mm_macc_pd(vfeps,Fp,Y);
 376             velec            = _mm_mul_pd(qq02,VV);
 377             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 378             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq02,FF),_mm_mul_pd(vftabscale,rinv02)));
 379
 380             /* Update potential sum for this i atom from the interaction with this j atom. */
 381             velecsum         = _mm_add_pd(velecsum,velec);
 382
 383             fscal            = felec;
 384
 385             /* Update vectorial force */
 386             fix0             = _mm_macc_pd(dx02,fscal,fix0);
 387             fiy0             = _mm_macc_pd(dy02,fscal,fiy0);
 388             fiz0             = _mm_macc_pd(dz02,fscal,fiz0);
 389
 390             fjx2             = _mm_macc_pd(dx02,fscal,fjx2);
 391             fjy2             = _mm_macc_pd(dy02,fscal,fjy2);
 392             fjz2             = _mm_macc_pd(dz02,fscal,fjz2);
 393
 394             /**************************
 395              * CALCULATE INTERACTIONS *
 396              **************************/
 397
 398             r10              = _mm_mul_pd(rsq10,rinv10);
 399
 400             /* Calculate table index by multiplying r with table scale and truncate to integer */
 401             rt               = _mm_mul_pd(r10,vftabscale);
 402             vfitab           = _mm_cvttpd_epi32(rt);
 403 #ifdef __XOP__
 404             vfeps            = _mm_frcz_pd(rt);
 405 #else
 406             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 407 #endif
 408             twovfeps         = _mm_add_pd(vfeps,vfeps);
 409             vfitab           = _mm_slli_epi32(vfitab,2);
 410
 411             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 412             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 413             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 414             GMX_MM_TRANSPOSE2_PD(Y,F);
 415             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 416             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 417             GMX_MM_TRANSPOSE2_PD(G,H);
 418             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 419             VV               = _mm_macc_pd(vfeps,Fp,Y);
 420             velec            = _mm_mul_pd(qq10,VV);
 421             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 422             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq10,FF),_mm_mul_pd(vftabscale,rinv10)));
 423
 424             /* Update potential sum for this i atom from the interaction with this j atom. */
 425             velecsum         = _mm_add_pd(velecsum,velec);
 426
 427             fscal            = felec;
 428
 429             /* Update vectorial force */
 430             fix1             = _mm_macc_pd(dx10,fscal,fix1);
 431             fiy1             = _mm_macc_pd(dy10,fscal,fiy1);
 432             fiz1             = _mm_macc_pd(dz10,fscal,fiz1);
 433
 434             fjx0             = _mm_macc_pd(dx10,fscal,fjx0);
 435             fjy0             = _mm_macc_pd(dy10,fscal,fjy0);
 436             fjz0             = _mm_macc_pd(dz10,fscal,fjz0);
 437
 438             /**************************
 439              * CALCULATE INTERACTIONS *
 440              **************************/
 441
 442             r11              = _mm_mul_pd(rsq11,rinv11);
 443
 444             /* Calculate table index by multiplying r with table scale and truncate to integer */
 445             rt               = _mm_mul_pd(r11,vftabscale);
 446             vfitab           = _mm_cvttpd_epi32(rt);
 447 #ifdef __XOP__
 448             vfeps            = _mm_frcz_pd(rt);
 449 #else
 450             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 451 #endif
 452             twovfeps         = _mm_add_pd(vfeps,vfeps);
 453             vfitab           = _mm_slli_epi32(vfitab,2);
 454
 455             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 456             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 457             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 458             GMX_MM_TRANSPOSE2_PD(Y,F);
 459             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 460             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 461             GMX_MM_TRANSPOSE2_PD(G,H);
 462             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 463             VV               = _mm_macc_pd(vfeps,Fp,Y);
 464             velec            = _mm_mul_pd(qq11,VV);
 465             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 466             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq11,FF),_mm_mul_pd(vftabscale,rinv11)));
 467
 468             /* Update potential sum for this i atom from the interaction with this j atom. */
 469             velecsum         = _mm_add_pd(velecsum,velec);
 470
 471             fscal            = felec;
 472
 473             /* Update vectorial force */
 474             fix1             = _mm_macc_pd(dx11,fscal,fix1);
 475             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
 476             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
 477
 478             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
 479             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
 480             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
 481
 482             /**************************
 483              * CALCULATE INTERACTIONS *
 484              **************************/
 485
 486             r12              = _mm_mul_pd(rsq12,rinv12);
 487
 488             /* Calculate table index by multiplying r with table scale and truncate to integer */
 489             rt               = _mm_mul_pd(r12,vftabscale);
 490             vfitab           = _mm_cvttpd_epi32(rt);
 491 #ifdef __XOP__
 492             vfeps            = _mm_frcz_pd(rt);
 493 #else
 494             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 495 #endif
 496             twovfeps         = _mm_add_pd(vfeps,vfeps);
 497             vfitab           = _mm_slli_epi32(vfitab,2);
 498
 499             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 500             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 501             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 502             GMX_MM_TRANSPOSE2_PD(Y,F);
 503             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 504             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 505             GMX_MM_TRANSPOSE2_PD(G,H);
 506             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 507             VV               = _mm_macc_pd(vfeps,Fp,Y);
 508             velec            = _mm_mul_pd(qq12,VV);
 509             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 510             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq12,FF),_mm_mul_pd(vftabscale,rinv12)));
 511
 512             /* Update potential sum for this i atom from the interaction with this j atom. */
 513             velecsum         = _mm_add_pd(velecsum,velec);
 514
 515             fscal            = felec;
 516
 517             /* Update vectorial force */
 518             fix1             = _mm_macc_pd(dx12,fscal,fix1);
 519             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
 520             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
 521
 522             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
 523             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
 524             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
 525
 526             /**************************
 527              * CALCULATE INTERACTIONS *
 528              **************************/
 529
 530             r20              = _mm_mul_pd(rsq20,rinv20);
 531
 532             /* Calculate table index by multiplying r with table scale and truncate to integer */
 533             rt               = _mm_mul_pd(r20,vftabscale);
 534             vfitab           = _mm_cvttpd_epi32(rt);
 535 #ifdef __XOP__
 536             vfeps            = _mm_frcz_pd(rt);
 537 #else
 538             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 539 #endif
 540             twovfeps         = _mm_add_pd(vfeps,vfeps);
 541             vfitab           = _mm_slli_epi32(vfitab,2);
 542
 543             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 544             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 545             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 546             GMX_MM_TRANSPOSE2_PD(Y,F);
 547             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 548             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 549             GMX_MM_TRANSPOSE2_PD(G,H);
 550             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 551             VV               = _mm_macc_pd(vfeps,Fp,Y);
 552             velec            = _mm_mul_pd(qq20,VV);
 553             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 554             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq20,FF),_mm_mul_pd(vftabscale,rinv20)));
 555
 556             /* Update potential sum for this i atom from the interaction with this j atom. */
 557             velecsum         = _mm_add_pd(velecsum,velec);
 558
 559             fscal            = felec;
 560
 561             /* Update vectorial force */
 562             fix2             = _mm_macc_pd(dx20,fscal,fix2);
 563             fiy2             = _mm_macc_pd(dy20,fscal,fiy2);
 564             fiz2             = _mm_macc_pd(dz20,fscal,fiz2);
 565
 566             fjx0             = _mm_macc_pd(dx20,fscal,fjx0);
 567             fjy0             = _mm_macc_pd(dy20,fscal,fjy0);
 568             fjz0             = _mm_macc_pd(dz20,fscal,fjz0);
 569
 570             /**************************
 571              * CALCULATE INTERACTIONS *
 572              **************************/
 573
 574             r21              = _mm_mul_pd(rsq21,rinv21);
 575
 576             /* Calculate table index by multiplying r with table scale and truncate to integer */
 577             rt               = _mm_mul_pd(r21,vftabscale);
 578             vfitab           = _mm_cvttpd_epi32(rt);
 579 #ifdef __XOP__
 580             vfeps            = _mm_frcz_pd(rt);
 581 #else
 582             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 583 #endif
 584             twovfeps         = _mm_add_pd(vfeps,vfeps);
 585             vfitab           = _mm_slli_epi32(vfitab,2);
 586
 587             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 588             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 589             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 590             GMX_MM_TRANSPOSE2_PD(Y,F);
 591             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 592             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 593             GMX_MM_TRANSPOSE2_PD(G,H);
 594             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 595             VV               = _mm_macc_pd(vfeps,Fp,Y);
 596             velec            = _mm_mul_pd(qq21,VV);
 597             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 598             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq21,FF),_mm_mul_pd(vftabscale,rinv21)));
 599
 600             /* Update potential sum for this i atom from the interaction with this j atom. */
 601             velecsum         = _mm_add_pd(velecsum,velec);
 602
 603             fscal            = felec;
 604
 605             /* Update vectorial force */
 606             fix2             = _mm_macc_pd(dx21,fscal,fix2);
 607             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
 608             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
 609
 610             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
 611             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
 612             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
 613
 614             /**************************
 615              * CALCULATE INTERACTIONS *
 616              **************************/
 617
 618             r22              = _mm_mul_pd(rsq22,rinv22);
 619
 620             /* Calculate table index by multiplying r with table scale and truncate to integer */
 621             rt               = _mm_mul_pd(r22,vftabscale);
 622             vfitab           = _mm_cvttpd_epi32(rt);
 623 #ifdef __XOP__
 624             vfeps            = _mm_frcz_pd(rt);
 625 #else
 626             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 627 #endif
 628             twovfeps         = _mm_add_pd(vfeps,vfeps);
 629             vfitab           = _mm_slli_epi32(vfitab,2);
 630
 631             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 632             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 633             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 634             GMX_MM_TRANSPOSE2_PD(Y,F);
 635             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 636             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 637             GMX_MM_TRANSPOSE2_PD(G,H);
 638             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 639             VV               = _mm_macc_pd(vfeps,Fp,Y);
 640             velec            = _mm_mul_pd(qq22,VV);
 641             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 642             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq22,FF),_mm_mul_pd(vftabscale,rinv22)));
 643
 644             /* Update potential sum for this i atom from the interaction with this j atom. */
 645             velecsum         = _mm_add_pd(velecsum,velec);
 646
 647             fscal            = felec;
 648
 649             /* Update vectorial force */
 650             fix2             = _mm_macc_pd(dx22,fscal,fix2);
 651             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
 652             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
 653
 654             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
 655             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
 656             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
 657
 658             gmx_mm_decrement_3rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
 659
 660             /* Inner loop uses 414 flops */
 661         }
 662
 663         if(jidx<j_index_end)
 664         {
 665
 666             jnrA             = jjnr[jidx];
 667             j_coord_offsetA  = DIM*jnrA;
 668
 669             /* load j atom coordinates */
 670             gmx_mm_load_3rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 671                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 672
 673             /* Calculate displacement vector */
 674             dx00             = _mm_sub_pd(ix0,jx0);
 675             dy00             = _mm_sub_pd(iy0,jy0);
 676             dz00             = _mm_sub_pd(iz0,jz0);
 677             dx01             = _mm_sub_pd(ix0,jx1);
 678             dy01             = _mm_sub_pd(iy0,jy1);
 679             dz01             = _mm_sub_pd(iz0,jz1);
 680             dx02             = _mm_sub_pd(ix0,jx2);
 681             dy02             = _mm_sub_pd(iy0,jy2);
 682             dz02             = _mm_sub_pd(iz0,jz2);
 683             dx10             = _mm_sub_pd(ix1,jx0);
 684             dy10             = _mm_sub_pd(iy1,jy0);
 685             dz10             = _mm_sub_pd(iz1,jz0);
 686             dx11             = _mm_sub_pd(ix1,jx1);
 687             dy11             = _mm_sub_pd(iy1,jy1);
 688             dz11             = _mm_sub_pd(iz1,jz1);
 689             dx12             = _mm_sub_pd(ix1,jx2);
 690             dy12             = _mm_sub_pd(iy1,jy2);
 691             dz12             = _mm_sub_pd(iz1,jz2);
 692             dx20             = _mm_sub_pd(ix2,jx0);
 693             dy20             = _mm_sub_pd(iy2,jy0);
 694             dz20             = _mm_sub_pd(iz2,jz0);
 695             dx21             = _mm_sub_pd(ix2,jx1);
 696             dy21             = _mm_sub_pd(iy2,jy1);
 697             dz21             = _mm_sub_pd(iz2,jz1);
 698             dx22             = _mm_sub_pd(ix2,jx2);
 699             dy22             = _mm_sub_pd(iy2,jy2);
 700             dz22             = _mm_sub_pd(iz2,jz2);
 701
 702             /* Calculate squared distance and things based on it */
 703             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 704             rsq01            = gmx_mm_calc_rsq_pd(dx01,dy01,dz01);
 705             rsq02            = gmx_mm_calc_rsq_pd(dx02,dy02,dz02);
 706             rsq10            = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
 707             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
 708             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
 709             rsq20            = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
 710             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
 711             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
 712
 713             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 714             rinv01           = gmx_mm_invsqrt_pd(rsq01);
 715             rinv02           = gmx_mm_invsqrt_pd(rsq02);
 716             rinv10           = gmx_mm_invsqrt_pd(rsq10);
 717             rinv11           = gmx_mm_invsqrt_pd(rsq11);
 718             rinv12           = gmx_mm_invsqrt_pd(rsq12);
 719             rinv20           = gmx_mm_invsqrt_pd(rsq20);
 720             rinv21           = gmx_mm_invsqrt_pd(rsq21);
 721             rinv22           = gmx_mm_invsqrt_pd(rsq22);
 722
 723             fjx0             = _mm_setzero_pd();
 724             fjy0             = _mm_setzero_pd();
 725             fjz0             = _mm_setzero_pd();
 726             fjx1             = _mm_setzero_pd();
 727             fjy1             = _mm_setzero_pd();
 728             fjz1             = _mm_setzero_pd();
 729             fjx2             = _mm_setzero_pd();
 730             fjy2             = _mm_setzero_pd();
 731             fjz2             = _mm_setzero_pd();
 732
 733             /**************************
 734              * CALCULATE INTERACTIONS *
 735              **************************/
 736
 737             r00              = _mm_mul_pd(rsq00,rinv00);
 738
 739             /* Calculate table index by multiplying r with table scale and truncate to integer */
 740             rt               = _mm_mul_pd(r00,vftabscale);
 741             vfitab           = _mm_cvttpd_epi32(rt);
 742 #ifdef __XOP__
 743             vfeps            = _mm_frcz_pd(rt);
 744 #else
 745             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 746 #endif
 747             twovfeps         = _mm_add_pd(vfeps,vfeps);
 748             vfitab           = _mm_slli_epi32(vfitab,2);
 749
 750             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 751             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 752             F                = _mm_setzero_pd();
 753             GMX_MM_TRANSPOSE2_PD(Y,F);
 754             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 755             H                = _mm_setzero_pd();
 756             GMX_MM_TRANSPOSE2_PD(G,H);
 757             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 758             VV               = _mm_macc_pd(vfeps,Fp,Y);
 759             velec            = _mm_mul_pd(qq00,VV);
 760             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 761             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq00,FF),_mm_mul_pd(vftabscale,rinv00)));
 762
 763             /* Update potential sum for this i atom from the interaction with this j atom. */
 764             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 765             velecsum         = _mm_add_pd(velecsum,velec);
 766
 767             fscal            = felec;
 768
 769             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 770
 771             /* Update vectorial force */
 772             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 773             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 774             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 775
 776             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
 777             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
 778             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
 779
 780             /**************************
 781              * CALCULATE INTERACTIONS *
 782              **************************/
 783
 784             r01              = _mm_mul_pd(rsq01,rinv01);
 785
 786             /* Calculate table index by multiplying r with table scale and truncate to integer */
 787             rt               = _mm_mul_pd(r01,vftabscale);
 788             vfitab           = _mm_cvttpd_epi32(rt);
 789 #ifdef __XOP__
 790             vfeps            = _mm_frcz_pd(rt);
 791 #else
 792             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 793 #endif
 794             twovfeps         = _mm_add_pd(vfeps,vfeps);
 795             vfitab           = _mm_slli_epi32(vfitab,2);
 796
 797             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 798             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 799             F                = _mm_setzero_pd();
 800             GMX_MM_TRANSPOSE2_PD(Y,F);
 801             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 802             H                = _mm_setzero_pd();
 803             GMX_MM_TRANSPOSE2_PD(G,H);
 804             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 805             VV               = _mm_macc_pd(vfeps,Fp,Y);
 806             velec            = _mm_mul_pd(qq01,VV);
 807             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 808             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq01,FF),_mm_mul_pd(vftabscale,rinv01)));
 809
 810             /* Update potential sum for this i atom from the interaction with this j atom. */
 811             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 812             velecsum         = _mm_add_pd(velecsum,velec);
 813
 814             fscal            = felec;
 815
 816             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 817
 818             /* Update vectorial force */
 819             fix0             = _mm_macc_pd(dx01,fscal,fix0);
 820             fiy0             = _mm_macc_pd(dy01,fscal,fiy0);
 821             fiz0             = _mm_macc_pd(dz01,fscal,fiz0);
 822
 823             fjx1             = _mm_macc_pd(dx01,fscal,fjx1);
 824             fjy1             = _mm_macc_pd(dy01,fscal,fjy1);
 825             fjz1             = _mm_macc_pd(dz01,fscal,fjz1);
 826
 827             /**************************
 828              * CALCULATE INTERACTIONS *
 829              **************************/
 830
 831             r02              = _mm_mul_pd(rsq02,rinv02);
 832
 833             /* Calculate table index by multiplying r with table scale and truncate to integer */
 834             rt               = _mm_mul_pd(r02,vftabscale);
 835             vfitab           = _mm_cvttpd_epi32(rt);
 836 #ifdef __XOP__
 837             vfeps            = _mm_frcz_pd(rt);
 838 #else
 839             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 840 #endif
 841             twovfeps         = _mm_add_pd(vfeps,vfeps);
 842             vfitab           = _mm_slli_epi32(vfitab,2);
 843
 844             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 845             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 846             F                = _mm_setzero_pd();
 847             GMX_MM_TRANSPOSE2_PD(Y,F);
 848             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 849             H                = _mm_setzero_pd();
 850             GMX_MM_TRANSPOSE2_PD(G,H);
 851             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 852             VV               = _mm_macc_pd(vfeps,Fp,Y);
 853             velec            = _mm_mul_pd(qq02,VV);
 854             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 855             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq02,FF),_mm_mul_pd(vftabscale,rinv02)));
 856
 857             /* Update potential sum for this i atom from the interaction with this j atom. */
 858             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 859             velecsum         = _mm_add_pd(velecsum,velec);
 860
 861             fscal            = felec;
 862
 863             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 864
 865             /* Update vectorial force */
 866             fix0             = _mm_macc_pd(dx02,fscal,fix0);
 867             fiy0             = _mm_macc_pd(dy02,fscal,fiy0);
 868             fiz0             = _mm_macc_pd(dz02,fscal,fiz0);
 869
 870             fjx2             = _mm_macc_pd(dx02,fscal,fjx2);
 871             fjy2             = _mm_macc_pd(dy02,fscal,fjy2);
 872             fjz2             = _mm_macc_pd(dz02,fscal,fjz2);
 873
 874             /**************************
 875              * CALCULATE INTERACTIONS *
 876              **************************/
 877
 878             r10              = _mm_mul_pd(rsq10,rinv10);
 879
 880             /* Calculate table index by multiplying r with table scale and truncate to integer */
 881             rt               = _mm_mul_pd(r10,vftabscale);
 882             vfitab           = _mm_cvttpd_epi32(rt);
 883 #ifdef __XOP__
 884             vfeps            = _mm_frcz_pd(rt);
 885 #else
 886             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 887 #endif
 888             twovfeps         = _mm_add_pd(vfeps,vfeps);
 889             vfitab           = _mm_slli_epi32(vfitab,2);
 890
 891             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 892             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 893             F                = _mm_setzero_pd();
 894             GMX_MM_TRANSPOSE2_PD(Y,F);
 895             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 896             H                = _mm_setzero_pd();
 897             GMX_MM_TRANSPOSE2_PD(G,H);
 898             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 899             VV               = _mm_macc_pd(vfeps,Fp,Y);
 900             velec            = _mm_mul_pd(qq10,VV);
 901             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 902             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq10,FF),_mm_mul_pd(vftabscale,rinv10)));
 903
 904             /* Update potential sum for this i atom from the interaction with this j atom. */
 905             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 906             velecsum         = _mm_add_pd(velecsum,velec);
 907
 908             fscal            = felec;
 909
 910             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 911
 912             /* Update vectorial force */
 913             fix1             = _mm_macc_pd(dx10,fscal,fix1);
 914             fiy1             = _mm_macc_pd(dy10,fscal,fiy1);
 915             fiz1             = _mm_macc_pd(dz10,fscal,fiz1);
 916
 917             fjx0             = _mm_macc_pd(dx10,fscal,fjx0);
 918             fjy0             = _mm_macc_pd(dy10,fscal,fjy0);
 919             fjz0             = _mm_macc_pd(dz10,fscal,fjz0);
 920
 921             /**************************
 922              * CALCULATE INTERACTIONS *
 923              **************************/
 924
 925             r11              = _mm_mul_pd(rsq11,rinv11);
 926
 927             /* Calculate table index by multiplying r with table scale and truncate to integer */
 928             rt               = _mm_mul_pd(r11,vftabscale);
 929             vfitab           = _mm_cvttpd_epi32(rt);
 930 #ifdef __XOP__
 931             vfeps            = _mm_frcz_pd(rt);
 932 #else
 933             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 934 #endif
 935             twovfeps         = _mm_add_pd(vfeps,vfeps);
 936             vfitab           = _mm_slli_epi32(vfitab,2);
 937
 938             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 939             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 940             F                = _mm_setzero_pd();
 941             GMX_MM_TRANSPOSE2_PD(Y,F);
 942             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 943             H                = _mm_setzero_pd();
 944             GMX_MM_TRANSPOSE2_PD(G,H);
 945             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 946             VV               = _mm_macc_pd(vfeps,Fp,Y);
 947             velec            = _mm_mul_pd(qq11,VV);
 948             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 949             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq11,FF),_mm_mul_pd(vftabscale,rinv11)));
 950
 951             /* Update potential sum for this i atom from the interaction with this j atom. */
 952             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 953             velecsum         = _mm_add_pd(velecsum,velec);
 954
 955             fscal            = felec;
 956
 957             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 958
 959             /* Update vectorial force */
 960             fix1             = _mm_macc_pd(dx11,fscal,fix1);
 961             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
 962             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
 963
 964             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
 965             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
 966             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
 967
 968             /**************************
 969              * CALCULATE INTERACTIONS *
 970              **************************/
 971
 972             r12              = _mm_mul_pd(rsq12,rinv12);
 973
 974             /* Calculate table index by multiplying r with table scale and truncate to integer */
 975             rt               = _mm_mul_pd(r12,vftabscale);
 976             vfitab           = _mm_cvttpd_epi32(rt);
 977 #ifdef __XOP__
 978             vfeps            = _mm_frcz_pd(rt);
 979 #else
 980             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 981 #endif
 982             twovfeps         = _mm_add_pd(vfeps,vfeps);
 983             vfitab           = _mm_slli_epi32(vfitab,2);
 984
 985             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 986             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 987             F                = _mm_setzero_pd();
 988             GMX_MM_TRANSPOSE2_PD(Y,F);
 989             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 990             H                = _mm_setzero_pd();
 991             GMX_MM_TRANSPOSE2_PD(G,H);
 992             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 993             VV               = _mm_macc_pd(vfeps,Fp,Y);
 994             velec            = _mm_mul_pd(qq12,VV);
 995             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 996             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq12,FF),_mm_mul_pd(vftabscale,rinv12)));
 997
 998             /* Update potential sum for this i atom from the interaction with this j atom. */
 999             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1000             velecsum         = _mm_add_pd(velecsum,velec);
1001
1002             fscal            = felec;
1003
1004             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1005
1006             /* Update vectorial force */
1007             fix1             = _mm_macc_pd(dx12,fscal,fix1);
1008             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
1009             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
1010
1011             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
1012             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
1013             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
1014
1015             /**************************
1016              * CALCULATE INTERACTIONS *
1017              **************************/
1018
1019             r20              = _mm_mul_pd(rsq20,rinv20);
1020
1021             /* Calculate table index by multiplying r with table scale and truncate to integer */
1022             rt               = _mm_mul_pd(r20,vftabscale);
1023             vfitab           = _mm_cvttpd_epi32(rt);
1024 #ifdef __XOP__
1025             vfeps            = _mm_frcz_pd(rt);
1026 #else
1027             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1028 #endif
1029             twovfeps         = _mm_add_pd(vfeps,vfeps);
1030             vfitab           = _mm_slli_epi32(vfitab,2);
1031
1032             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1033             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1034             F                = _mm_setzero_pd();
1035             GMX_MM_TRANSPOSE2_PD(Y,F);
1036             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1037             H                = _mm_setzero_pd();
1038             GMX_MM_TRANSPOSE2_PD(G,H);
1039             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1040             VV               = _mm_macc_pd(vfeps,Fp,Y);
1041             velec            = _mm_mul_pd(qq20,VV);
1042             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1043             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq20,FF),_mm_mul_pd(vftabscale,rinv20)));
1044
1045             /* Update potential sum for this i atom from the interaction with this j atom. */
1046             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1047             velecsum         = _mm_add_pd(velecsum,velec);
1048
1049             fscal            = felec;
1050
1051             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1052
1053             /* Update vectorial force */
1054             fix2             = _mm_macc_pd(dx20,fscal,fix2);
1055             fiy2             = _mm_macc_pd(dy20,fscal,fiy2);
1056             fiz2             = _mm_macc_pd(dz20,fscal,fiz2);
1057
1058             fjx0             = _mm_macc_pd(dx20,fscal,fjx0);
1059             fjy0             = _mm_macc_pd(dy20,fscal,fjy0);
1060             fjz0             = _mm_macc_pd(dz20,fscal,fjz0);
1061
1062             /**************************
1063              * CALCULATE INTERACTIONS *
1064              **************************/
1065
1066             r21              = _mm_mul_pd(rsq21,rinv21);
1067
1068             /* Calculate table index by multiplying r with table scale and truncate to integer */
1069             rt               = _mm_mul_pd(r21,vftabscale);
1070             vfitab           = _mm_cvttpd_epi32(rt);
1071 #ifdef __XOP__
1072             vfeps            = _mm_frcz_pd(rt);
1073 #else
1074             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1075 #endif
1076             twovfeps         = _mm_add_pd(vfeps,vfeps);
1077             vfitab           = _mm_slli_epi32(vfitab,2);
1078
1079             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1080             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1081             F                = _mm_setzero_pd();
1082             GMX_MM_TRANSPOSE2_PD(Y,F);
1083             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1084             H                = _mm_setzero_pd();
1085             GMX_MM_TRANSPOSE2_PD(G,H);
1086             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1087             VV               = _mm_macc_pd(vfeps,Fp,Y);
1088             velec            = _mm_mul_pd(qq21,VV);
1089             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1090             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq21,FF),_mm_mul_pd(vftabscale,rinv21)));
1091
1092             /* Update potential sum for this i atom from the interaction with this j atom. */
1093             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1094             velecsum         = _mm_add_pd(velecsum,velec);
1095
1096             fscal            = felec;
1097
1098             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1099
1100             /* Update vectorial force */
1101             fix2             = _mm_macc_pd(dx21,fscal,fix2);
1102             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
1103             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
1104
1105             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
1106             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
1107             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
1108
1109             /**************************
1110              * CALCULATE INTERACTIONS *
1111              **************************/
1112
1113             r22              = _mm_mul_pd(rsq22,rinv22);
1114
1115             /* Calculate table index by multiplying r with table scale and truncate to integer */
1116             rt               = _mm_mul_pd(r22,vftabscale);
1117             vfitab           = _mm_cvttpd_epi32(rt);
1118 #ifdef __XOP__
1119             vfeps            = _mm_frcz_pd(rt);
1120 #else
1121             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1122 #endif
1123             twovfeps         = _mm_add_pd(vfeps,vfeps);
1124             vfitab           = _mm_slli_epi32(vfitab,2);
1125
1126             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1127             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1128             F                = _mm_setzero_pd();
1129             GMX_MM_TRANSPOSE2_PD(Y,F);
1130             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1131             H                = _mm_setzero_pd();
1132             GMX_MM_TRANSPOSE2_PD(G,H);
1133             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1134             VV               = _mm_macc_pd(vfeps,Fp,Y);
1135             velec            = _mm_mul_pd(qq22,VV);
1136             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1137             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq22,FF),_mm_mul_pd(vftabscale,rinv22)));
1138
1139             /* Update potential sum for this i atom from the interaction with this j atom. */
1140             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1141             velecsum         = _mm_add_pd(velecsum,velec);
1142
1143             fscal            = felec;
1144
1145             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1146
1147             /* Update vectorial force */
1148             fix2             = _mm_macc_pd(dx22,fscal,fix2);
1149             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
1150             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
1151
1152             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
1153             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
1154             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
1155
1156             gmx_mm_decrement_3rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1157
1158             /* Inner loop uses 414 flops */
1159         }
1160
1161         /* End of innermost loop */
1162
1163         gmx_mm_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
1164                                               f+i_coord_offset,fshift+i_shift_offset);
1165
1166         ggid                        = gid[iidx];
1167         /* Update potential energies */
1168         gmx_mm_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1169
1170         /* Increment number of inner iterations */
1171         inneriter                  += j_index_end - j_index_start;
1172
1173         /* Outer loop uses 19 flops */
1174     }
1175
1176     /* Increment number of outer iterations */
1177     outeriter        += nri;
1178
1179     /* Update outer/inner flops */
1180
1181     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W3W3_VF,outeriter*19 + inneriter*414);
1182 }
1183 /*
1184  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwNone_GeomW3W3_F_avx_128_fma_double
1185  * Electrostatics interaction: CubicSplineTable
1186  * VdW interaction:            None
1187  * Geometry:                   Water3-Water3
1188  * Calculate force/pot:        Force
1189  */
1190 void
1191 nb_kernel_ElecCSTab_VdwNone_GeomW3W3_F_avx_128_fma_double
1192                     (t_nblist                    * gmx_restrict       nlist,
1193                      rvec                        * gmx_restrict          xx,
1194                      rvec                        * gmx_restrict          ff,
1195                      t_forcerec                  * gmx_restrict          fr,
1196                      t_mdatoms                   * gmx_restrict     mdatoms,
1197                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1198                      t_nrnb                      * gmx_restrict        nrnb)
1199 {
1200     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1201      * just 0 for non-waters.
1202      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
1203      * jnr indices corresponding to data put in the four positions in the SIMD register.
1204      */
1205     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
1206     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1207     int              jnrA,jnrB;
1208     int              j_coord_offsetA,j_coord_offsetB;
1209     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
1210     real             rcutoff_scalar;
1211     real             *shiftvec,*fshift,*x,*f;
1212     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1213     int              vdwioffset0;
1214     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1215     int              vdwioffset1;
1216     __m128d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1217     int              vdwioffset2;
1218     __m128d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1219     int              vdwjidx0A,vdwjidx0B;
1220     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1221     int              vdwjidx1A,vdwjidx1B;
1222     __m128d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1223     int              vdwjidx2A,vdwjidx2B;
1224     __m128d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1225     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1226     __m128d          dx01,dy01,dz01,rsq01,rinv01,rinvsq01,r01,qq01,c6_01,c12_01;
1227     __m128d          dx02,dy02,dz02,rsq02,rinv02,rinvsq02,r02,qq02,c6_02,c12_02;
1228     __m128d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
1229     __m128d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1230     __m128d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1231     __m128d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
1232     __m128d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1233     __m128d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1234     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
1235     real             *charge;
1236     __m128i          vfitab;
1237     __m128i          ifour       = _mm_set1_epi32(4);
1238     __m128d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF,twovfeps;
1239     real             *vftab;
1240     __m128d          dummy_mask,cutoff_mask;
1241     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
1242     __m128d          one     = _mm_set1_pd(1.0);
1243     __m128d          two     = _mm_set1_pd(2.0);
1244     x                = xx[0];
1245     f                = ff[0];
1246
1247     nri              = nlist->nri;
1248     iinr             = nlist->iinr;
1249     jindex           = nlist->jindex;
1250     jjnr             = nlist->jjnr;
1251     shiftidx         = nlist->shift;
1252     gid              = nlist->gid;
1253     shiftvec         = fr->shift_vec[0];
1254     fshift           = fr->fshift[0];
1255     facel            = _mm_set1_pd(fr->epsfac);
1256     charge           = mdatoms->chargeA;
1257
1258     vftab            = kernel_data->table_elec->data;
1259     vftabscale       = _mm_set1_pd(kernel_data->table_elec->scale);
1260
1261     /* Setup water-specific parameters */
1262     inr              = nlist->iinr[0];
1263     iq0              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+0]));
1264     iq1              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
1265     iq2              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
1266
1267     jq0              = _mm_set1_pd(charge[inr+0]);
1268     jq1              = _mm_set1_pd(charge[inr+1]);
1269     jq2              = _mm_set1_pd(charge[inr+2]);
1270     qq00             = _mm_mul_pd(iq0,jq0);
1271     qq01             = _mm_mul_pd(iq0,jq1);
1272     qq02             = _mm_mul_pd(iq0,jq2);
1273     qq10             = _mm_mul_pd(iq1,jq0);
1274     qq11             = _mm_mul_pd(iq1,jq1);
1275     qq12             = _mm_mul_pd(iq1,jq2);
1276     qq20             = _mm_mul_pd(iq2,jq0);
1277     qq21             = _mm_mul_pd(iq2,jq1);
1278     qq22             = _mm_mul_pd(iq2,jq2);
1279
1280     /* Avoid stupid compiler warnings */
1281     jnrA = jnrB = 0;
1282     j_coord_offsetA = 0;
1283     j_coord_offsetB = 0;
1284
1285     outeriter        = 0;
1286     inneriter        = 0;
1287
1288     /* Start outer loop over neighborlists */
1289     for(iidx=0; iidx<nri; iidx++)
1290     {
1291         /* Load shift vector for this list */
1292         i_shift_offset   = DIM*shiftidx[iidx];
1293
1294         /* Load limits for loop over neighbors */
1295         j_index_start    = jindex[iidx];
1296         j_index_end      = jindex[iidx+1];
1297
1298         /* Get outer coordinate index */
1299         inr              = iinr[iidx];
1300         i_coord_offset   = DIM*inr;
1301
1302         /* Load i particle coords and add shift vector */
1303         gmx_mm_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1304                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
1305
1306         fix0             = _mm_setzero_pd();
1307         fiy0             = _mm_setzero_pd();
1308         fiz0             = _mm_setzero_pd();
1309         fix1             = _mm_setzero_pd();
1310         fiy1             = _mm_setzero_pd();
1311         fiz1             = _mm_setzero_pd();
1312         fix2             = _mm_setzero_pd();
1313         fiy2             = _mm_setzero_pd();
1314         fiz2             = _mm_setzero_pd();
1315
1316         /* Start inner kernel loop */
1317         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
1318         {
1319
1320             /* Get j neighbor index, and coordinate index */
1321             jnrA             = jjnr[jidx];
1322             jnrB             = jjnr[jidx+1];
1323             j_coord_offsetA  = DIM*jnrA;
1324             j_coord_offsetB  = DIM*jnrB;
1325
1326             /* load j atom coordinates */
1327             gmx_mm_load_3rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1328                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
1329
1330             /* Calculate displacement vector */
1331             dx00             = _mm_sub_pd(ix0,jx0);
1332             dy00             = _mm_sub_pd(iy0,jy0);
1333             dz00             = _mm_sub_pd(iz0,jz0);
1334             dx01             = _mm_sub_pd(ix0,jx1);
1335             dy01             = _mm_sub_pd(iy0,jy1);
1336             dz01             = _mm_sub_pd(iz0,jz1);
1337             dx02             = _mm_sub_pd(ix0,jx2);
1338             dy02             = _mm_sub_pd(iy0,jy2);
1339             dz02             = _mm_sub_pd(iz0,jz2);
1340             dx10             = _mm_sub_pd(ix1,jx0);
1341             dy10             = _mm_sub_pd(iy1,jy0);
1342             dz10             = _mm_sub_pd(iz1,jz0);
1343             dx11             = _mm_sub_pd(ix1,jx1);
1344             dy11             = _mm_sub_pd(iy1,jy1);
1345             dz11             = _mm_sub_pd(iz1,jz1);
1346             dx12             = _mm_sub_pd(ix1,jx2);
1347             dy12             = _mm_sub_pd(iy1,jy2);
1348             dz12             = _mm_sub_pd(iz1,jz2);
1349             dx20             = _mm_sub_pd(ix2,jx0);
1350             dy20             = _mm_sub_pd(iy2,jy0);
1351             dz20             = _mm_sub_pd(iz2,jz0);
1352             dx21             = _mm_sub_pd(ix2,jx1);
1353             dy21             = _mm_sub_pd(iy2,jy1);
1354             dz21             = _mm_sub_pd(iz2,jz1);
1355             dx22             = _mm_sub_pd(ix2,jx2);
1356             dy22             = _mm_sub_pd(iy2,jy2);
1357             dz22             = _mm_sub_pd(iz2,jz2);
1358
1359             /* Calculate squared distance and things based on it */
1360             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1361             rsq01            = gmx_mm_calc_rsq_pd(dx01,dy01,dz01);
1362             rsq02            = gmx_mm_calc_rsq_pd(dx02,dy02,dz02);
1363             rsq10            = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
1364             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
1365             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
1366             rsq20            = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
1367             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
1368             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
1369
1370             rinv00           = gmx_mm_invsqrt_pd(rsq00);
1371             rinv01           = gmx_mm_invsqrt_pd(rsq01);
1372             rinv02           = gmx_mm_invsqrt_pd(rsq02);
1373             rinv10           = gmx_mm_invsqrt_pd(rsq10);
1374             rinv11           = gmx_mm_invsqrt_pd(rsq11);
1375             rinv12           = gmx_mm_invsqrt_pd(rsq12);
1376             rinv20           = gmx_mm_invsqrt_pd(rsq20);
1377             rinv21           = gmx_mm_invsqrt_pd(rsq21);
1378             rinv22           = gmx_mm_invsqrt_pd(rsq22);
1379
1380             fjx0             = _mm_setzero_pd();
1381             fjy0             = _mm_setzero_pd();
1382             fjz0             = _mm_setzero_pd();
1383             fjx1             = _mm_setzero_pd();
1384             fjy1             = _mm_setzero_pd();
1385             fjz1             = _mm_setzero_pd();
1386             fjx2             = _mm_setzero_pd();
1387             fjy2             = _mm_setzero_pd();
1388             fjz2             = _mm_setzero_pd();
1389
1390             /**************************
1391              * CALCULATE INTERACTIONS *
1392              **************************/
1393
1394             r00              = _mm_mul_pd(rsq00,rinv00);
1395
1396             /* Calculate table index by multiplying r with table scale and truncate to integer */
1397             rt               = _mm_mul_pd(r00,vftabscale);
1398             vfitab           = _mm_cvttpd_epi32(rt);
1399 #ifdef __XOP__
1400             vfeps            = _mm_frcz_pd(rt);
1401 #else
1402             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1403 #endif
1404             twovfeps         = _mm_add_pd(vfeps,vfeps);
1405             vfitab           = _mm_slli_epi32(vfitab,2);
1406
1407             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1408             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1409             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1410             GMX_MM_TRANSPOSE2_PD(Y,F);
1411             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1412             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1413             GMX_MM_TRANSPOSE2_PD(G,H);
1414             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1415             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1416             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq00,FF),_mm_mul_pd(vftabscale,rinv00)));
1417
1418             fscal            = felec;
1419
1420             /* Update vectorial force */
1421             fix0             = _mm_macc_pd(dx00,fscal,fix0);
1422             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
1423             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
1424
1425             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
1426             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
1427             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
1428
1429             /**************************
1430              * CALCULATE INTERACTIONS *
1431              **************************/
1432
1433             r01              = _mm_mul_pd(rsq01,rinv01);
1434
1435             /* Calculate table index by multiplying r with table scale and truncate to integer */
1436             rt               = _mm_mul_pd(r01,vftabscale);
1437             vfitab           = _mm_cvttpd_epi32(rt);
1438 #ifdef __XOP__
1439             vfeps            = _mm_frcz_pd(rt);
1440 #else
1441             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1442 #endif
1443             twovfeps         = _mm_add_pd(vfeps,vfeps);
1444             vfitab           = _mm_slli_epi32(vfitab,2);
1445
1446             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1447             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1448             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1449             GMX_MM_TRANSPOSE2_PD(Y,F);
1450             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1451             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1452             GMX_MM_TRANSPOSE2_PD(G,H);
1453             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1454             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1455             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq01,FF),_mm_mul_pd(vftabscale,rinv01)));
1456
1457             fscal            = felec;
1458
1459             /* Update vectorial force */
1460             fix0             = _mm_macc_pd(dx01,fscal,fix0);
1461             fiy0             = _mm_macc_pd(dy01,fscal,fiy0);
1462             fiz0             = _mm_macc_pd(dz01,fscal,fiz0);
1463
1464             fjx1             = _mm_macc_pd(dx01,fscal,fjx1);
1465             fjy1             = _mm_macc_pd(dy01,fscal,fjy1);
1466             fjz1             = _mm_macc_pd(dz01,fscal,fjz1);
1467
1468             /**************************
1469              * CALCULATE INTERACTIONS *
1470              **************************/
1471
1472             r02              = _mm_mul_pd(rsq02,rinv02);
1473
1474             /* Calculate table index by multiplying r with table scale and truncate to integer */
1475             rt               = _mm_mul_pd(r02,vftabscale);
1476             vfitab           = _mm_cvttpd_epi32(rt);
1477 #ifdef __XOP__
1478             vfeps            = _mm_frcz_pd(rt);
1479 #else
1480             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1481 #endif
1482             twovfeps         = _mm_add_pd(vfeps,vfeps);
1483             vfitab           = _mm_slli_epi32(vfitab,2);
1484
1485             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1486             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1487             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1488             GMX_MM_TRANSPOSE2_PD(Y,F);
1489             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1490             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1491             GMX_MM_TRANSPOSE2_PD(G,H);
1492             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1493             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1494             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq02,FF),_mm_mul_pd(vftabscale,rinv02)));
1495
1496             fscal            = felec;
1497
1498             /* Update vectorial force */
1499             fix0             = _mm_macc_pd(dx02,fscal,fix0);
1500             fiy0             = _mm_macc_pd(dy02,fscal,fiy0);
1501             fiz0             = _mm_macc_pd(dz02,fscal,fiz0);
1502
1503             fjx2             = _mm_macc_pd(dx02,fscal,fjx2);
1504             fjy2             = _mm_macc_pd(dy02,fscal,fjy2);
1505             fjz2             = _mm_macc_pd(dz02,fscal,fjz2);
1506
1507             /**************************
1508              * CALCULATE INTERACTIONS *
1509              **************************/
1510
1511             r10              = _mm_mul_pd(rsq10,rinv10);
1512
1513             /* Calculate table index by multiplying r with table scale and truncate to integer */
1514             rt               = _mm_mul_pd(r10,vftabscale);
1515             vfitab           = _mm_cvttpd_epi32(rt);
1516 #ifdef __XOP__
1517             vfeps            = _mm_frcz_pd(rt);
1518 #else
1519             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1520 #endif
1521             twovfeps         = _mm_add_pd(vfeps,vfeps);
1522             vfitab           = _mm_slli_epi32(vfitab,2);
1523
1524             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1525             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1526             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1527             GMX_MM_TRANSPOSE2_PD(Y,F);
1528             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1529             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1530             GMX_MM_TRANSPOSE2_PD(G,H);
1531             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1532             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1533             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq10,FF),_mm_mul_pd(vftabscale,rinv10)));
1534
1535             fscal            = felec;
1536
1537             /* Update vectorial force */
1538             fix1             = _mm_macc_pd(dx10,fscal,fix1);
1539             fiy1             = _mm_macc_pd(dy10,fscal,fiy1);
1540             fiz1             = _mm_macc_pd(dz10,fscal,fiz1);
1541
1542             fjx0             = _mm_macc_pd(dx10,fscal,fjx0);
1543             fjy0             = _mm_macc_pd(dy10,fscal,fjy0);
1544             fjz0             = _mm_macc_pd(dz10,fscal,fjz0);
1545
1546             /**************************
1547              * CALCULATE INTERACTIONS *
1548              **************************/
1549
1550             r11              = _mm_mul_pd(rsq11,rinv11);
1551
1552             /* Calculate table index by multiplying r with table scale and truncate to integer */
1553             rt               = _mm_mul_pd(r11,vftabscale);
1554             vfitab           = _mm_cvttpd_epi32(rt);
1555 #ifdef __XOP__
1556             vfeps            = _mm_frcz_pd(rt);
1557 #else
1558             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1559 #endif
1560             twovfeps         = _mm_add_pd(vfeps,vfeps);
1561             vfitab           = _mm_slli_epi32(vfitab,2);
1562
1563             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1564             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1565             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1566             GMX_MM_TRANSPOSE2_PD(Y,F);
1567             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1568             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1569             GMX_MM_TRANSPOSE2_PD(G,H);
1570             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1571             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1572             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq11,FF),_mm_mul_pd(vftabscale,rinv11)));
1573
1574             fscal            = felec;
1575
1576             /* Update vectorial force */
1577             fix1             = _mm_macc_pd(dx11,fscal,fix1);
1578             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
1579             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
1580
1581             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
1582             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
1583             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
1584
1585             /**************************
1586              * CALCULATE INTERACTIONS *
1587              **************************/
1588
1589             r12              = _mm_mul_pd(rsq12,rinv12);
1590
1591             /* Calculate table index by multiplying r with table scale and truncate to integer */
1592             rt               = _mm_mul_pd(r12,vftabscale);
1593             vfitab           = _mm_cvttpd_epi32(rt);
1594 #ifdef __XOP__
1595             vfeps            = _mm_frcz_pd(rt);
1596 #else
1597             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1598 #endif
1599             twovfeps         = _mm_add_pd(vfeps,vfeps);
1600             vfitab           = _mm_slli_epi32(vfitab,2);
1601
1602             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1603             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1604             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1605             GMX_MM_TRANSPOSE2_PD(Y,F);
1606             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1607             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1608             GMX_MM_TRANSPOSE2_PD(G,H);
1609             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1610             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1611             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq12,FF),_mm_mul_pd(vftabscale,rinv12)));
1612
1613             fscal            = felec;
1614
1615             /* Update vectorial force */
1616             fix1             = _mm_macc_pd(dx12,fscal,fix1);
1617             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
1618             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
1619
1620             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
1621             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
1622             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
1623
1624             /**************************
1625              * CALCULATE INTERACTIONS *
1626              **************************/
1627
1628             r20              = _mm_mul_pd(rsq20,rinv20);
1629
1630             /* Calculate table index by multiplying r with table scale and truncate to integer */
1631             rt               = _mm_mul_pd(r20,vftabscale);
1632             vfitab           = _mm_cvttpd_epi32(rt);
1633 #ifdef __XOP__
1634             vfeps            = _mm_frcz_pd(rt);
1635 #else
1636             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1637 #endif
1638             twovfeps         = _mm_add_pd(vfeps,vfeps);
1639             vfitab           = _mm_slli_epi32(vfitab,2);
1640
1641             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1642             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1643             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1644             GMX_MM_TRANSPOSE2_PD(Y,F);
1645             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1646             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1647             GMX_MM_TRANSPOSE2_PD(G,H);
1648             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1649             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1650             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq20,FF),_mm_mul_pd(vftabscale,rinv20)));
1651
1652             fscal            = felec;
1653
1654             /* Update vectorial force */
1655             fix2             = _mm_macc_pd(dx20,fscal,fix2);
1656             fiy2             = _mm_macc_pd(dy20,fscal,fiy2);
1657             fiz2             = _mm_macc_pd(dz20,fscal,fiz2);
1658
1659             fjx0             = _mm_macc_pd(dx20,fscal,fjx0);
1660             fjy0             = _mm_macc_pd(dy20,fscal,fjy0);
1661             fjz0             = _mm_macc_pd(dz20,fscal,fjz0);
1662
1663             /**************************
1664              * CALCULATE INTERACTIONS *
1665              **************************/
1666
1667             r21              = _mm_mul_pd(rsq21,rinv21);
1668
1669             /* Calculate table index by multiplying r with table scale and truncate to integer */
1670             rt               = _mm_mul_pd(r21,vftabscale);
1671             vfitab           = _mm_cvttpd_epi32(rt);
1672 #ifdef __XOP__
1673             vfeps            = _mm_frcz_pd(rt);
1674 #else
1675             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1676 #endif
1677             twovfeps         = _mm_add_pd(vfeps,vfeps);
1678             vfitab           = _mm_slli_epi32(vfitab,2);
1679
1680             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1681             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1682             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1683             GMX_MM_TRANSPOSE2_PD(Y,F);
1684             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1685             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1686             GMX_MM_TRANSPOSE2_PD(G,H);
1687             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1688             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1689             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq21,FF),_mm_mul_pd(vftabscale,rinv21)));
1690
1691             fscal            = felec;
1692
1693             /* Update vectorial force */
1694             fix2             = _mm_macc_pd(dx21,fscal,fix2);
1695             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
1696             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
1697
1698             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
1699             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
1700             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
1701
1702             /**************************
1703              * CALCULATE INTERACTIONS *
1704              **************************/
1705
1706             r22              = _mm_mul_pd(rsq22,rinv22);
1707
1708             /* Calculate table index by multiplying r with table scale and truncate to integer */
1709             rt               = _mm_mul_pd(r22,vftabscale);
1710             vfitab           = _mm_cvttpd_epi32(rt);
1711 #ifdef __XOP__
1712             vfeps            = _mm_frcz_pd(rt);
1713 #else
1714             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1715 #endif
1716             twovfeps         = _mm_add_pd(vfeps,vfeps);
1717             vfitab           = _mm_slli_epi32(vfitab,2);
1718
1719             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1720             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1721             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1722             GMX_MM_TRANSPOSE2_PD(Y,F);
1723             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1724             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1725             GMX_MM_TRANSPOSE2_PD(G,H);
1726             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1727             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1728             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq22,FF),_mm_mul_pd(vftabscale,rinv22)));
1729
1730             fscal            = felec;
1731
1732             /* Update vectorial force */
1733             fix2             = _mm_macc_pd(dx22,fscal,fix2);
1734             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
1735             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
1736
1737             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
1738             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
1739             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
1740
1741             gmx_mm_decrement_3rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1742
1743             /* Inner loop uses 378 flops */
1744         }
1745
1746         if(jidx<j_index_end)
1747         {
1748
1749             jnrA             = jjnr[jidx];
1750             j_coord_offsetA  = DIM*jnrA;
1751
1752             /* load j atom coordinates */
1753             gmx_mm_load_3rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
1754                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
1755
1756             /* Calculate displacement vector */
1757             dx00             = _mm_sub_pd(ix0,jx0);
1758             dy00             = _mm_sub_pd(iy0,jy0);
1759             dz00             = _mm_sub_pd(iz0,jz0);
1760             dx01             = _mm_sub_pd(ix0,jx1);
1761             dy01             = _mm_sub_pd(iy0,jy1);
1762             dz01             = _mm_sub_pd(iz0,jz1);
1763             dx02             = _mm_sub_pd(ix0,jx2);
1764             dy02             = _mm_sub_pd(iy0,jy2);
1765             dz02             = _mm_sub_pd(iz0,jz2);
1766             dx10             = _mm_sub_pd(ix1,jx0);
1767             dy10             = _mm_sub_pd(iy1,jy0);
1768             dz10             = _mm_sub_pd(iz1,jz0);
1769             dx11             = _mm_sub_pd(ix1,jx1);
1770             dy11             = _mm_sub_pd(iy1,jy1);
1771             dz11             = _mm_sub_pd(iz1,jz1);
1772             dx12             = _mm_sub_pd(ix1,jx2);
1773             dy12             = _mm_sub_pd(iy1,jy2);
1774             dz12             = _mm_sub_pd(iz1,jz2);
1775             dx20             = _mm_sub_pd(ix2,jx0);
1776             dy20             = _mm_sub_pd(iy2,jy0);
1777             dz20             = _mm_sub_pd(iz2,jz0);
1778             dx21             = _mm_sub_pd(ix2,jx1);
1779             dy21             = _mm_sub_pd(iy2,jy1);
1780             dz21             = _mm_sub_pd(iz2,jz1);
1781             dx22             = _mm_sub_pd(ix2,jx2);
1782             dy22             = _mm_sub_pd(iy2,jy2);
1783             dz22             = _mm_sub_pd(iz2,jz2);
1784
1785             /* Calculate squared distance and things based on it */
1786             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1787             rsq01            = gmx_mm_calc_rsq_pd(dx01,dy01,dz01);
1788             rsq02            = gmx_mm_calc_rsq_pd(dx02,dy02,dz02);
1789             rsq10            = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
1790             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
1791             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
1792             rsq20            = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
1793             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
1794             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
1795
1796             rinv00           = gmx_mm_invsqrt_pd(rsq00);
1797             rinv01           = gmx_mm_invsqrt_pd(rsq01);
1798             rinv02           = gmx_mm_invsqrt_pd(rsq02);
1799             rinv10           = gmx_mm_invsqrt_pd(rsq10);
1800             rinv11           = gmx_mm_invsqrt_pd(rsq11);
1801             rinv12           = gmx_mm_invsqrt_pd(rsq12);
1802             rinv20           = gmx_mm_invsqrt_pd(rsq20);
1803             rinv21           = gmx_mm_invsqrt_pd(rsq21);
1804             rinv22           = gmx_mm_invsqrt_pd(rsq22);
1805
1806             fjx0             = _mm_setzero_pd();
1807             fjy0             = _mm_setzero_pd();
1808             fjz0             = _mm_setzero_pd();
1809             fjx1             = _mm_setzero_pd();
1810             fjy1             = _mm_setzero_pd();
1811             fjz1             = _mm_setzero_pd();
1812             fjx2             = _mm_setzero_pd();
1813             fjy2             = _mm_setzero_pd();
1814             fjz2             = _mm_setzero_pd();
1815
1816             /**************************
1817              * CALCULATE INTERACTIONS *
1818              **************************/
1819
1820             r00              = _mm_mul_pd(rsq00,rinv00);
1821
1822             /* Calculate table index by multiplying r with table scale and truncate to integer */
1823             rt               = _mm_mul_pd(r00,vftabscale);
1824             vfitab           = _mm_cvttpd_epi32(rt);
1825 #ifdef __XOP__
1826             vfeps            = _mm_frcz_pd(rt);
1827 #else
1828             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1829 #endif
1830             twovfeps         = _mm_add_pd(vfeps,vfeps);
1831             vfitab           = _mm_slli_epi32(vfitab,2);
1832
1833             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1834             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1835             F                = _mm_setzero_pd();
1836             GMX_MM_TRANSPOSE2_PD(Y,F);
1837             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1838             H                = _mm_setzero_pd();
1839             GMX_MM_TRANSPOSE2_PD(G,H);
1840             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1841             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1842             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq00,FF),_mm_mul_pd(vftabscale,rinv00)));
1843
1844             fscal            = felec;
1845
1846             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1847
1848             /* Update vectorial force */
1849             fix0             = _mm_macc_pd(dx00,fscal,fix0);
1850             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
1851             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
1852
1853             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
1854             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
1855             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
1856
1857             /**************************
1858              * CALCULATE INTERACTIONS *
1859              **************************/
1860
1861             r01              = _mm_mul_pd(rsq01,rinv01);
1862
1863             /* Calculate table index by multiplying r with table scale and truncate to integer */
1864             rt               = _mm_mul_pd(r01,vftabscale);
1865             vfitab           = _mm_cvttpd_epi32(rt);
1866 #ifdef __XOP__
1867             vfeps            = _mm_frcz_pd(rt);
1868 #else
1869             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1870 #endif
1871             twovfeps         = _mm_add_pd(vfeps,vfeps);
1872             vfitab           = _mm_slli_epi32(vfitab,2);
1873
1874             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1875             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1876             F                = _mm_setzero_pd();
1877             GMX_MM_TRANSPOSE2_PD(Y,F);
1878             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1879             H                = _mm_setzero_pd();
1880             GMX_MM_TRANSPOSE2_PD(G,H);
1881             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1882             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1883             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq01,FF),_mm_mul_pd(vftabscale,rinv01)));
1884
1885             fscal            = felec;
1886
1887             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1888
1889             /* Update vectorial force */
1890             fix0             = _mm_macc_pd(dx01,fscal,fix0);
1891             fiy0             = _mm_macc_pd(dy01,fscal,fiy0);
1892             fiz0             = _mm_macc_pd(dz01,fscal,fiz0);
1893
1894             fjx1             = _mm_macc_pd(dx01,fscal,fjx1);
1895             fjy1             = _mm_macc_pd(dy01,fscal,fjy1);
1896             fjz1             = _mm_macc_pd(dz01,fscal,fjz1);
1897
1898             /**************************
1899              * CALCULATE INTERACTIONS *
1900              **************************/
1901
1902             r02              = _mm_mul_pd(rsq02,rinv02);
1903
1904             /* Calculate table index by multiplying r with table scale and truncate to integer */
1905             rt               = _mm_mul_pd(r02,vftabscale);
1906             vfitab           = _mm_cvttpd_epi32(rt);
1907 #ifdef __XOP__
1908             vfeps            = _mm_frcz_pd(rt);
1909 #else
1910             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1911 #endif
1912             twovfeps         = _mm_add_pd(vfeps,vfeps);
1913             vfitab           = _mm_slli_epi32(vfitab,2);
1914
1915             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1916             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1917             F                = _mm_setzero_pd();
1918             GMX_MM_TRANSPOSE2_PD(Y,F);
1919             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1920             H                = _mm_setzero_pd();
1921             GMX_MM_TRANSPOSE2_PD(G,H);
1922             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1923             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1924             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq02,FF),_mm_mul_pd(vftabscale,rinv02)));
1925
1926             fscal            = felec;
1927
1928             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1929
1930             /* Update vectorial force */
1931             fix0             = _mm_macc_pd(dx02,fscal,fix0);
1932             fiy0             = _mm_macc_pd(dy02,fscal,fiy0);
1933             fiz0             = _mm_macc_pd(dz02,fscal,fiz0);
1934
1935             fjx2             = _mm_macc_pd(dx02,fscal,fjx2);
1936             fjy2             = _mm_macc_pd(dy02,fscal,fjy2);
1937             fjz2             = _mm_macc_pd(dz02,fscal,fjz2);
1938
1939             /**************************
1940              * CALCULATE INTERACTIONS *
1941              **************************/
1942
1943             r10              = _mm_mul_pd(rsq10,rinv10);
1944
1945             /* Calculate table index by multiplying r with table scale and truncate to integer */
1946             rt               = _mm_mul_pd(r10,vftabscale);
1947             vfitab           = _mm_cvttpd_epi32(rt);
1948 #ifdef __XOP__
1949             vfeps            = _mm_frcz_pd(rt);
1950 #else
1951             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1952 #endif
1953             twovfeps         = _mm_add_pd(vfeps,vfeps);
1954             vfitab           = _mm_slli_epi32(vfitab,2);
1955
1956             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1957             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1958             F                = _mm_setzero_pd();
1959             GMX_MM_TRANSPOSE2_PD(Y,F);
1960             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1961             H                = _mm_setzero_pd();
1962             GMX_MM_TRANSPOSE2_PD(G,H);
1963             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
1964             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
1965             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq10,FF),_mm_mul_pd(vftabscale,rinv10)));
1966
1967             fscal            = felec;
1968
1969             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1970
1971             /* Update vectorial force */
1972             fix1             = _mm_macc_pd(dx10,fscal,fix1);
1973             fiy1             = _mm_macc_pd(dy10,fscal,fiy1);
1974             fiz1             = _mm_macc_pd(dz10,fscal,fiz1);
1975
1976             fjx0             = _mm_macc_pd(dx10,fscal,fjx0);
1977             fjy0             = _mm_macc_pd(dy10,fscal,fjy0);
1978             fjz0             = _mm_macc_pd(dz10,fscal,fjz0);
1979
1980             /**************************
1981              * CALCULATE INTERACTIONS *
1982              **************************/
1983
1984             r11              = _mm_mul_pd(rsq11,rinv11);
1985
1986             /* Calculate table index by multiplying r with table scale and truncate to integer */
1987             rt               = _mm_mul_pd(r11,vftabscale);
1988             vfitab           = _mm_cvttpd_epi32(rt);
1989 #ifdef __XOP__
1990             vfeps            = _mm_frcz_pd(rt);
1991 #else
1992             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1993 #endif
1994             twovfeps         = _mm_add_pd(vfeps,vfeps);
1995             vfitab           = _mm_slli_epi32(vfitab,2);
1996
1997             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1998             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1999             F                = _mm_setzero_pd();
2000             GMX_MM_TRANSPOSE2_PD(Y,F);
2001             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
2002             H                = _mm_setzero_pd();
2003             GMX_MM_TRANSPOSE2_PD(G,H);
2004             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
2005             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
2006             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq11,FF),_mm_mul_pd(vftabscale,rinv11)));
2007
2008             fscal            = felec;
2009
2010             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2011
2012             /* Update vectorial force */
2013             fix1             = _mm_macc_pd(dx11,fscal,fix1);
2014             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
2015             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
2016
2017             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
2018             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
2019             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
2020
2021             /**************************
2022              * CALCULATE INTERACTIONS *
2023              **************************/
2024
2025             r12              = _mm_mul_pd(rsq12,rinv12);
2026
2027             /* Calculate table index by multiplying r with table scale and truncate to integer */
2028             rt               = _mm_mul_pd(r12,vftabscale);
2029             vfitab           = _mm_cvttpd_epi32(rt);
2030 #ifdef __XOP__
2031             vfeps            = _mm_frcz_pd(rt);
2032 #else
2033             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
2034 #endif
2035             twovfeps         = _mm_add_pd(vfeps,vfeps);
2036             vfitab           = _mm_slli_epi32(vfitab,2);
2037
2038             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2039             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
2040             F                = _mm_setzero_pd();
2041             GMX_MM_TRANSPOSE2_PD(Y,F);
2042             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
2043             H                = _mm_setzero_pd();
2044             GMX_MM_TRANSPOSE2_PD(G,H);
2045             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
2046             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
2047             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq12,FF),_mm_mul_pd(vftabscale,rinv12)));
2048
2049             fscal            = felec;
2050
2051             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2052
2053             /* Update vectorial force */
2054             fix1             = _mm_macc_pd(dx12,fscal,fix1);
2055             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
2056             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
2057
2058             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
2059             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
2060             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
2061
2062             /**************************
2063              * CALCULATE INTERACTIONS *
2064              **************************/
2065
2066             r20              = _mm_mul_pd(rsq20,rinv20);
2067
2068             /* Calculate table index by multiplying r with table scale and truncate to integer */
2069             rt               = _mm_mul_pd(r20,vftabscale);
2070             vfitab           = _mm_cvttpd_epi32(rt);
2071 #ifdef __XOP__
2072             vfeps            = _mm_frcz_pd(rt);
2073 #else
2074             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
2075 #endif
2076             twovfeps         = _mm_add_pd(vfeps,vfeps);
2077             vfitab           = _mm_slli_epi32(vfitab,2);
2078
2079             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2080             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
2081             F                = _mm_setzero_pd();
2082             GMX_MM_TRANSPOSE2_PD(Y,F);
2083             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
2084             H                = _mm_setzero_pd();
2085             GMX_MM_TRANSPOSE2_PD(G,H);
2086             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
2087             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
2088             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq20,FF),_mm_mul_pd(vftabscale,rinv20)));
2089
2090             fscal            = felec;
2091
2092             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2093
2094             /* Update vectorial force */
2095             fix2             = _mm_macc_pd(dx20,fscal,fix2);
2096             fiy2             = _mm_macc_pd(dy20,fscal,fiy2);
2097             fiz2             = _mm_macc_pd(dz20,fscal,fiz2);
2098
2099             fjx0             = _mm_macc_pd(dx20,fscal,fjx0);
2100             fjy0             = _mm_macc_pd(dy20,fscal,fjy0);
2101             fjz0             = _mm_macc_pd(dz20,fscal,fjz0);
2102
2103             /**************************
2104              * CALCULATE INTERACTIONS *
2105              **************************/
2106
2107             r21              = _mm_mul_pd(rsq21,rinv21);
2108
2109             /* Calculate table index by multiplying r with table scale and truncate to integer */
2110             rt               = _mm_mul_pd(r21,vftabscale);
2111             vfitab           = _mm_cvttpd_epi32(rt);
2112 #ifdef __XOP__
2113             vfeps            = _mm_frcz_pd(rt);
2114 #else
2115             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
2116 #endif
2117             twovfeps         = _mm_add_pd(vfeps,vfeps);
2118             vfitab           = _mm_slli_epi32(vfitab,2);
2119
2120             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2121             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
2122             F                = _mm_setzero_pd();
2123             GMX_MM_TRANSPOSE2_PD(Y,F);
2124             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
2125             H                = _mm_setzero_pd();
2126             GMX_MM_TRANSPOSE2_PD(G,H);
2127             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
2128             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
2129             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq21,FF),_mm_mul_pd(vftabscale,rinv21)));
2130
2131             fscal            = felec;
2132
2133             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2134
2135             /* Update vectorial force */
2136             fix2             = _mm_macc_pd(dx21,fscal,fix2);
2137             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
2138             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
2139
2140             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
2141             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
2142             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
2143
2144             /**************************
2145              * CALCULATE INTERACTIONS *
2146              **************************/
2147
2148             r22              = _mm_mul_pd(rsq22,rinv22);
2149
2150             /* Calculate table index by multiplying r with table scale and truncate to integer */
2151             rt               = _mm_mul_pd(r22,vftabscale);
2152             vfitab           = _mm_cvttpd_epi32(rt);
2153 #ifdef __XOP__
2154             vfeps            = _mm_frcz_pd(rt);
2155 #else
2156             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
2157 #endif
2158             twovfeps         = _mm_add_pd(vfeps,vfeps);
2159             vfitab           = _mm_slli_epi32(vfitab,2);
2160
2161             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2162             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
2163             F                = _mm_setzero_pd();
2164             GMX_MM_TRANSPOSE2_PD(Y,F);
2165             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
2166             H                = _mm_setzero_pd();
2167             GMX_MM_TRANSPOSE2_PD(G,H);
2168             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
2169             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
2170             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq22,FF),_mm_mul_pd(vftabscale,rinv22)));
2171
2172             fscal            = felec;
2173
2174             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2175
2176             /* Update vectorial force */
2177             fix2             = _mm_macc_pd(dx22,fscal,fix2);
2178             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
2179             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
2180
2181             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
2182             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
2183             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
2184
2185             gmx_mm_decrement_3rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
2186
2187             /* Inner loop uses 378 flops */
2188         }
2189
2190         /* End of innermost loop */
2191
2192         gmx_mm_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
2193                                               f+i_coord_offset,fshift+i_shift_offset);
2194
2195         /* Increment number of inner iterations */
2196         inneriter                  += j_index_end - j_index_start;
2197
2198         /* Outer loop uses 18 flops */
2199     }
2200
2201     /* Increment number of outer iterations */
2202     outeriter        += nri;
2203
2204     /* Update outer/inner flops */
2205
2206     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W3W3_F,outeriter*18 + inneriter*378);
2207 }