src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_double/nb_kernel_ElecEw_VdwCSTab_GeomW4W4_avx_128_fma_double.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS avx_128_fma_double kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "gromacs/legacyheaders/types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "gromacs/legacyheaders/nrnb.h"
  46
  47 #include "gromacs/simd/math_x86_avx_128_fma_double.h"
  48 #include "kernelutil_x86_avx_128_fma_double.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwCSTab_GeomW4W4_VF_avx_128_fma_double
  52  * Electrostatics interaction: Ewald
  53  * VdW interaction:            CubicSplineTable
  54  * Geometry:                   Water4-Water4
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecEw_VdwCSTab_GeomW4W4_VF_avx_128_fma_double
  59                     (t_nblist                    * gmx_restrict       nlist,
  60                      rvec                        * gmx_restrict          xx,
  61                      rvec                        * gmx_restrict          ff,
  62                      t_forcerec                  * gmx_restrict          fr,
  63                      t_mdatoms                   * gmx_restrict     mdatoms,
  64                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  65                      t_nrnb                      * gmx_restrict        nrnb)
  66 {
  67     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  68      * just 0 for non-waters.
  69      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
  70      * jnr indices corresponding to data put in the four positions in the SIMD register.
  71      */
  72     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  73     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  74     int              jnrA,jnrB;
  75     int              j_coord_offsetA,j_coord_offsetB;
  76     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  77     real             rcutoff_scalar;
  78     real             *shiftvec,*fshift,*x,*f;
  79     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  80     int              vdwioffset0;
  81     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  82     int              vdwioffset1;
  83     __m128d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  84     int              vdwioffset2;
  85     __m128d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  86     int              vdwioffset3;
  87     __m128d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  88     int              vdwjidx0A,vdwjidx0B;
  89     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  90     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     int              vdwjidx3A,vdwjidx3B;
  95     __m128d          jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
  96     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  97     __m128d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
  98     __m128d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
  99     __m128d          dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
 100     __m128d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
 101     __m128d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
 102     __m128d          dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
 103     __m128d          dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
 104     __m128d          dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
 105     __m128d          dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
 106     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
 107     real             *charge;
 108     int              nvdwtype;
 109     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 110     int              *vdwtype;
 111     real             *vdwparam;
 112     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
 113     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
 114     __m128i          vfitab;
 115     __m128i          ifour       = _mm_set1_epi32(4);
 116     __m128d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF,twovfeps;
 117     real             *vftab;
 118     __m128i          ewitab;
 119     __m128d          ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 120     real             *ewtab;
 121     __m128d          dummy_mask,cutoff_mask;
 122     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
 123     __m128d          one     = _mm_set1_pd(1.0);
 124     __m128d          two     = _mm_set1_pd(2.0);
 125     x                = xx[0];
 126     f                = ff[0];
 127
 128     nri              = nlist->nri;
 129     iinr             = nlist->iinr;
 130     jindex           = nlist->jindex;
 131     jjnr             = nlist->jjnr;
 132     shiftidx         = nlist->shift;
 133     gid              = nlist->gid;
 134     shiftvec         = fr->shift_vec[0];
 135     fshift           = fr->fshift[0];
 136     facel            = _mm_set1_pd(fr->epsfac);
 137     charge           = mdatoms->chargeA;
 138     nvdwtype         = fr->ntype;
 139     vdwparam         = fr->nbfp;
 140     vdwtype          = mdatoms->typeA;
 141
 142     vftab            = kernel_data->table_vdw->data;
 143     vftabscale       = _mm_set1_pd(kernel_data->table_vdw->scale);
 144
 145     sh_ewald         = _mm_set1_pd(fr->ic->sh_ewald);
 146     ewtab            = fr->ic->tabq_coul_FDV0;
 147     ewtabscale       = _mm_set1_pd(fr->ic->tabq_scale);
 148     ewtabhalfspace   = _mm_set1_pd(0.5/fr->ic->tabq_scale);
 149
 150     /* Setup water-specific parameters */
 151     inr              = nlist->iinr[0];
 152     iq1              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
 153     iq2              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
 154     iq3              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
 155     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 156
 157     jq1              = _mm_set1_pd(charge[inr+1]);
 158     jq2              = _mm_set1_pd(charge[inr+2]);
 159     jq3              = _mm_set1_pd(charge[inr+3]);
 160     vdwjidx0A        = 2*vdwtype[inr+0];
 161     c6_00            = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A]);
 162     c12_00           = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A+1]);
 163     qq11             = _mm_mul_pd(iq1,jq1);
 164     qq12             = _mm_mul_pd(iq1,jq2);
 165     qq13             = _mm_mul_pd(iq1,jq3);
 166     qq21             = _mm_mul_pd(iq2,jq1);
 167     qq22             = _mm_mul_pd(iq2,jq2);
 168     qq23             = _mm_mul_pd(iq2,jq3);
 169     qq31             = _mm_mul_pd(iq3,jq1);
 170     qq32             = _mm_mul_pd(iq3,jq2);
 171     qq33             = _mm_mul_pd(iq3,jq3);
 172
 173     /* Avoid stupid compiler warnings */
 174     jnrA = jnrB = 0;
 175     j_coord_offsetA = 0;
 176     j_coord_offsetB = 0;
 177
 178     outeriter        = 0;
 179     inneriter        = 0;
 180
 181     /* Start outer loop over neighborlists */
 182     for(iidx=0; iidx<nri; iidx++)
 183     {
 184         /* Load shift vector for this list */
 185         i_shift_offset   = DIM*shiftidx[iidx];
 186
 187         /* Load limits for loop over neighbors */
 188         j_index_start    = jindex[iidx];
 189         j_index_end      = jindex[iidx+1];
 190
 191         /* Get outer coordinate index */
 192         inr              = iinr[iidx];
 193         i_coord_offset   = DIM*inr;
 194
 195         /* Load i particle coords and add shift vector */
 196         gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 197                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 198
 199         fix0             = _mm_setzero_pd();
 200         fiy0             = _mm_setzero_pd();
 201         fiz0             = _mm_setzero_pd();
 202         fix1             = _mm_setzero_pd();
 203         fiy1             = _mm_setzero_pd();
 204         fiz1             = _mm_setzero_pd();
 205         fix2             = _mm_setzero_pd();
 206         fiy2             = _mm_setzero_pd();
 207         fiz2             = _mm_setzero_pd();
 208         fix3             = _mm_setzero_pd();
 209         fiy3             = _mm_setzero_pd();
 210         fiz3             = _mm_setzero_pd();
 211
 212         /* Reset potential sums */
 213         velecsum         = _mm_setzero_pd();
 214         vvdwsum          = _mm_setzero_pd();
 215
 216         /* Start inner kernel loop */
 217         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
 218         {
 219
 220             /* Get j neighbor index, and coordinate index */
 221             jnrA             = jjnr[jidx];
 222             jnrB             = jjnr[jidx+1];
 223             j_coord_offsetA  = DIM*jnrA;
 224             j_coord_offsetB  = DIM*jnrB;
 225
 226             /* load j atom coordinates */
 227             gmx_mm_load_4rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 228                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 229                                               &jy2,&jz2,&jx3,&jy3,&jz3);
 230
 231             /* Calculate displacement vector */
 232             dx00             = _mm_sub_pd(ix0,jx0);
 233             dy00             = _mm_sub_pd(iy0,jy0);
 234             dz00             = _mm_sub_pd(iz0,jz0);
 235             dx11             = _mm_sub_pd(ix1,jx1);
 236             dy11             = _mm_sub_pd(iy1,jy1);
 237             dz11             = _mm_sub_pd(iz1,jz1);
 238             dx12             = _mm_sub_pd(ix1,jx2);
 239             dy12             = _mm_sub_pd(iy1,jy2);
 240             dz12             = _mm_sub_pd(iz1,jz2);
 241             dx13             = _mm_sub_pd(ix1,jx3);
 242             dy13             = _mm_sub_pd(iy1,jy3);
 243             dz13             = _mm_sub_pd(iz1,jz3);
 244             dx21             = _mm_sub_pd(ix2,jx1);
 245             dy21             = _mm_sub_pd(iy2,jy1);
 246             dz21             = _mm_sub_pd(iz2,jz1);
 247             dx22             = _mm_sub_pd(ix2,jx2);
 248             dy22             = _mm_sub_pd(iy2,jy2);
 249             dz22             = _mm_sub_pd(iz2,jz2);
 250             dx23             = _mm_sub_pd(ix2,jx3);
 251             dy23             = _mm_sub_pd(iy2,jy3);
 252             dz23             = _mm_sub_pd(iz2,jz3);
 253             dx31             = _mm_sub_pd(ix3,jx1);
 254             dy31             = _mm_sub_pd(iy3,jy1);
 255             dz31             = _mm_sub_pd(iz3,jz1);
 256             dx32             = _mm_sub_pd(ix3,jx2);
 257             dy32             = _mm_sub_pd(iy3,jy2);
 258             dz32             = _mm_sub_pd(iz3,jz2);
 259             dx33             = _mm_sub_pd(ix3,jx3);
 260             dy33             = _mm_sub_pd(iy3,jy3);
 261             dz33             = _mm_sub_pd(iz3,jz3);
 262
 263             /* Calculate squared distance and things based on it */
 264             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 265             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
 266             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
 267             rsq13            = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
 268             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
 269             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
 270             rsq23            = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
 271             rsq31            = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
 272             rsq32            = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
 273             rsq33            = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
 274
 275             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 276             rinv11           = gmx_mm_invsqrt_pd(rsq11);
 277             rinv12           = gmx_mm_invsqrt_pd(rsq12);
 278             rinv13           = gmx_mm_invsqrt_pd(rsq13);
 279             rinv21           = gmx_mm_invsqrt_pd(rsq21);
 280             rinv22           = gmx_mm_invsqrt_pd(rsq22);
 281             rinv23           = gmx_mm_invsqrt_pd(rsq23);
 282             rinv31           = gmx_mm_invsqrt_pd(rsq31);
 283             rinv32           = gmx_mm_invsqrt_pd(rsq32);
 284             rinv33           = gmx_mm_invsqrt_pd(rsq33);
 285
 286             rinvsq11         = _mm_mul_pd(rinv11,rinv11);
 287             rinvsq12         = _mm_mul_pd(rinv12,rinv12);
 288             rinvsq13         = _mm_mul_pd(rinv13,rinv13);
 289             rinvsq21         = _mm_mul_pd(rinv21,rinv21);
 290             rinvsq22         = _mm_mul_pd(rinv22,rinv22);
 291             rinvsq23         = _mm_mul_pd(rinv23,rinv23);
 292             rinvsq31         = _mm_mul_pd(rinv31,rinv31);
 293             rinvsq32         = _mm_mul_pd(rinv32,rinv32);
 294             rinvsq33         = _mm_mul_pd(rinv33,rinv33);
 295
 296             fjx0             = _mm_setzero_pd();
 297             fjy0             = _mm_setzero_pd();
 298             fjz0             = _mm_setzero_pd();
 299             fjx1             = _mm_setzero_pd();
 300             fjy1             = _mm_setzero_pd();
 301             fjz1             = _mm_setzero_pd();
 302             fjx2             = _mm_setzero_pd();
 303             fjy2             = _mm_setzero_pd();
 304             fjz2             = _mm_setzero_pd();
 305             fjx3             = _mm_setzero_pd();
 306             fjy3             = _mm_setzero_pd();
 307             fjz3             = _mm_setzero_pd();
 308
 309             /**************************
 310              * CALCULATE INTERACTIONS *
 311              **************************/
 312
 313             r00              = _mm_mul_pd(rsq00,rinv00);
 314
 315             /* Calculate table index by multiplying r with table scale and truncate to integer */
 316             rt               = _mm_mul_pd(r00,vftabscale);
 317             vfitab           = _mm_cvttpd_epi32(rt);
 318 #ifdef __XOP__
 319             vfeps            = _mm_frcz_pd(rt);
 320 #else
 321             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 322 #endif
 323             twovfeps         = _mm_add_pd(vfeps,vfeps);
 324             vfitab           = _mm_slli_epi32(vfitab,3);
 325
 326             /* CUBIC SPLINE TABLE DISPERSION */
 327             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 328             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 329             GMX_MM_TRANSPOSE2_PD(Y,F);
 330             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 331             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 332             GMX_MM_TRANSPOSE2_PD(G,H);
 333             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 334             VV               = _mm_macc_pd(vfeps,Fp,Y);
 335             vvdw6            = _mm_mul_pd(c6_00,VV);
 336             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 337             fvdw6            = _mm_mul_pd(c6_00,FF);
 338
 339             /* CUBIC SPLINE TABLE REPULSION */
 340             vfitab           = _mm_add_epi32(vfitab,ifour);
 341             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 342             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 343             GMX_MM_TRANSPOSE2_PD(Y,F);
 344             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 345             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 346             GMX_MM_TRANSPOSE2_PD(G,H);
 347             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 348             VV               = _mm_macc_pd(vfeps,Fp,Y);
 349             vvdw12           = _mm_mul_pd(c12_00,VV);
 350             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 351             fvdw12           = _mm_mul_pd(c12_00,FF);
 352             vvdw             = _mm_add_pd(vvdw12,vvdw6);
 353             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
 354
 355             /* Update potential sum for this i atom from the interaction with this j atom. */
 356             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 357
 358             fscal            = fvdw;
 359
 360             /* Update vectorial force */
 361             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 362             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 363             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 364
 365             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
 366             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
 367             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
 368
 369             /**************************
 370              * CALCULATE INTERACTIONS *
 371              **************************/
 372
 373             r11              = _mm_mul_pd(rsq11,rinv11);
 374
 375             /* EWALD ELECTROSTATICS */
 376
 377             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 378             ewrt             = _mm_mul_pd(r11,ewtabscale);
 379             ewitab           = _mm_cvttpd_epi32(ewrt);
 380 #ifdef __XOP__
 381             eweps            = _mm_frcz_pd(ewrt);
 382 #else
 383             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 384 #endif
 385             twoeweps         = _mm_add_pd(eweps,eweps);
 386             ewitab           = _mm_slli_epi32(ewitab,2);
 387             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 388             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 389             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 390             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 391             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 392             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 393             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 394             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 395             velec            = _mm_mul_pd(qq11,_mm_sub_pd(rinv11,velec));
 396             felec            = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
 397
 398             /* Update potential sum for this i atom from the interaction with this j atom. */
 399             velecsum         = _mm_add_pd(velecsum,velec);
 400
 401             fscal            = felec;
 402
 403             /* Update vectorial force */
 404             fix1             = _mm_macc_pd(dx11,fscal,fix1);
 405             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
 406             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
 407
 408             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
 409             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
 410             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
 411
 412             /**************************
 413              * CALCULATE INTERACTIONS *
 414              **************************/
 415
 416             r12              = _mm_mul_pd(rsq12,rinv12);
 417
 418             /* EWALD ELECTROSTATICS */
 419
 420             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 421             ewrt             = _mm_mul_pd(r12,ewtabscale);
 422             ewitab           = _mm_cvttpd_epi32(ewrt);
 423 #ifdef __XOP__
 424             eweps            = _mm_frcz_pd(ewrt);
 425 #else
 426             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 427 #endif
 428             twoeweps         = _mm_add_pd(eweps,eweps);
 429             ewitab           = _mm_slli_epi32(ewitab,2);
 430             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 431             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 432             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 433             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 434             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 435             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 436             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 437             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 438             velec            = _mm_mul_pd(qq12,_mm_sub_pd(rinv12,velec));
 439             felec            = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
 440
 441             /* Update potential sum for this i atom from the interaction with this j atom. */
 442             velecsum         = _mm_add_pd(velecsum,velec);
 443
 444             fscal            = felec;
 445
 446             /* Update vectorial force */
 447             fix1             = _mm_macc_pd(dx12,fscal,fix1);
 448             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
 449             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
 450
 451             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
 452             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
 453             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
 454
 455             /**************************
 456              * CALCULATE INTERACTIONS *
 457              **************************/
 458
 459             r13              = _mm_mul_pd(rsq13,rinv13);
 460
 461             /* EWALD ELECTROSTATICS */
 462
 463             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 464             ewrt             = _mm_mul_pd(r13,ewtabscale);
 465             ewitab           = _mm_cvttpd_epi32(ewrt);
 466 #ifdef __XOP__
 467             eweps            = _mm_frcz_pd(ewrt);
 468 #else
 469             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 470 #endif
 471             twoeweps         = _mm_add_pd(eweps,eweps);
 472             ewitab           = _mm_slli_epi32(ewitab,2);
 473             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 474             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 475             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 476             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 477             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 478             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 479             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 480             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 481             velec            = _mm_mul_pd(qq13,_mm_sub_pd(rinv13,velec));
 482             felec            = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
 483
 484             /* Update potential sum for this i atom from the interaction with this j atom. */
 485             velecsum         = _mm_add_pd(velecsum,velec);
 486
 487             fscal            = felec;
 488
 489             /* Update vectorial force */
 490             fix1             = _mm_macc_pd(dx13,fscal,fix1);
 491             fiy1             = _mm_macc_pd(dy13,fscal,fiy1);
 492             fiz1             = _mm_macc_pd(dz13,fscal,fiz1);
 493
 494             fjx3             = _mm_macc_pd(dx13,fscal,fjx3);
 495             fjy3             = _mm_macc_pd(dy13,fscal,fjy3);
 496             fjz3             = _mm_macc_pd(dz13,fscal,fjz3);
 497
 498             /**************************
 499              * CALCULATE INTERACTIONS *
 500              **************************/
 501
 502             r21              = _mm_mul_pd(rsq21,rinv21);
 503
 504             /* EWALD ELECTROSTATICS */
 505
 506             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 507             ewrt             = _mm_mul_pd(r21,ewtabscale);
 508             ewitab           = _mm_cvttpd_epi32(ewrt);
 509 #ifdef __XOP__
 510             eweps            = _mm_frcz_pd(ewrt);
 511 #else
 512             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 513 #endif
 514             twoeweps         = _mm_add_pd(eweps,eweps);
 515             ewitab           = _mm_slli_epi32(ewitab,2);
 516             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 517             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 518             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 519             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 520             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 521             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 522             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 523             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 524             velec            = _mm_mul_pd(qq21,_mm_sub_pd(rinv21,velec));
 525             felec            = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
 526
 527             /* Update potential sum for this i atom from the interaction with this j atom. */
 528             velecsum         = _mm_add_pd(velecsum,velec);
 529
 530             fscal            = felec;
 531
 532             /* Update vectorial force */
 533             fix2             = _mm_macc_pd(dx21,fscal,fix2);
 534             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
 535             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
 536
 537             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
 538             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
 539             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
 540
 541             /**************************
 542              * CALCULATE INTERACTIONS *
 543              **************************/
 544
 545             r22              = _mm_mul_pd(rsq22,rinv22);
 546
 547             /* EWALD ELECTROSTATICS */
 548
 549             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 550             ewrt             = _mm_mul_pd(r22,ewtabscale);
 551             ewitab           = _mm_cvttpd_epi32(ewrt);
 552 #ifdef __XOP__
 553             eweps            = _mm_frcz_pd(ewrt);
 554 #else
 555             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 556 #endif
 557             twoeweps         = _mm_add_pd(eweps,eweps);
 558             ewitab           = _mm_slli_epi32(ewitab,2);
 559             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 560             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 561             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 562             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 563             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 564             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 565             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 566             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 567             velec            = _mm_mul_pd(qq22,_mm_sub_pd(rinv22,velec));
 568             felec            = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
 569
 570             /* Update potential sum for this i atom from the interaction with this j atom. */
 571             velecsum         = _mm_add_pd(velecsum,velec);
 572
 573             fscal            = felec;
 574
 575             /* Update vectorial force */
 576             fix2             = _mm_macc_pd(dx22,fscal,fix2);
 577             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
 578             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
 579
 580             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
 581             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
 582             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
 583
 584             /**************************
 585              * CALCULATE INTERACTIONS *
 586              **************************/
 587
 588             r23              = _mm_mul_pd(rsq23,rinv23);
 589
 590             /* EWALD ELECTROSTATICS */
 591
 592             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 593             ewrt             = _mm_mul_pd(r23,ewtabscale);
 594             ewitab           = _mm_cvttpd_epi32(ewrt);
 595 #ifdef __XOP__
 596             eweps            = _mm_frcz_pd(ewrt);
 597 #else
 598             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 599 #endif
 600             twoeweps         = _mm_add_pd(eweps,eweps);
 601             ewitab           = _mm_slli_epi32(ewitab,2);
 602             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 603             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 604             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 605             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 606             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 607             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 608             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 609             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 610             velec            = _mm_mul_pd(qq23,_mm_sub_pd(rinv23,velec));
 611             felec            = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
 612
 613             /* Update potential sum for this i atom from the interaction with this j atom. */
 614             velecsum         = _mm_add_pd(velecsum,velec);
 615
 616             fscal            = felec;
 617
 618             /* Update vectorial force */
 619             fix2             = _mm_macc_pd(dx23,fscal,fix2);
 620             fiy2             = _mm_macc_pd(dy23,fscal,fiy2);
 621             fiz2             = _mm_macc_pd(dz23,fscal,fiz2);
 622
 623             fjx3             = _mm_macc_pd(dx23,fscal,fjx3);
 624             fjy3             = _mm_macc_pd(dy23,fscal,fjy3);
 625             fjz3             = _mm_macc_pd(dz23,fscal,fjz3);
 626
 627             /**************************
 628              * CALCULATE INTERACTIONS *
 629              **************************/
 630
 631             r31              = _mm_mul_pd(rsq31,rinv31);
 632
 633             /* EWALD ELECTROSTATICS */
 634
 635             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 636             ewrt             = _mm_mul_pd(r31,ewtabscale);
 637             ewitab           = _mm_cvttpd_epi32(ewrt);
 638 #ifdef __XOP__
 639             eweps            = _mm_frcz_pd(ewrt);
 640 #else
 641             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 642 #endif
 643             twoeweps         = _mm_add_pd(eweps,eweps);
 644             ewitab           = _mm_slli_epi32(ewitab,2);
 645             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 646             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 647             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 648             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 649             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 650             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 651             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 652             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 653             velec            = _mm_mul_pd(qq31,_mm_sub_pd(rinv31,velec));
 654             felec            = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
 655
 656             /* Update potential sum for this i atom from the interaction with this j atom. */
 657             velecsum         = _mm_add_pd(velecsum,velec);
 658
 659             fscal            = felec;
 660
 661             /* Update vectorial force */
 662             fix3             = _mm_macc_pd(dx31,fscal,fix3);
 663             fiy3             = _mm_macc_pd(dy31,fscal,fiy3);
 664             fiz3             = _mm_macc_pd(dz31,fscal,fiz3);
 665
 666             fjx1             = _mm_macc_pd(dx31,fscal,fjx1);
 667             fjy1             = _mm_macc_pd(dy31,fscal,fjy1);
 668             fjz1             = _mm_macc_pd(dz31,fscal,fjz1);
 669
 670             /**************************
 671              * CALCULATE INTERACTIONS *
 672              **************************/
 673
 674             r32              = _mm_mul_pd(rsq32,rinv32);
 675
 676             /* EWALD ELECTROSTATICS */
 677
 678             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 679             ewrt             = _mm_mul_pd(r32,ewtabscale);
 680             ewitab           = _mm_cvttpd_epi32(ewrt);
 681 #ifdef __XOP__
 682             eweps            = _mm_frcz_pd(ewrt);
 683 #else
 684             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 685 #endif
 686             twoeweps         = _mm_add_pd(eweps,eweps);
 687             ewitab           = _mm_slli_epi32(ewitab,2);
 688             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 689             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 690             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 691             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 692             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 693             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 694             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 695             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 696             velec            = _mm_mul_pd(qq32,_mm_sub_pd(rinv32,velec));
 697             felec            = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
 698
 699             /* Update potential sum for this i atom from the interaction with this j atom. */
 700             velecsum         = _mm_add_pd(velecsum,velec);
 701
 702             fscal            = felec;
 703
 704             /* Update vectorial force */
 705             fix3             = _mm_macc_pd(dx32,fscal,fix3);
 706             fiy3             = _mm_macc_pd(dy32,fscal,fiy3);
 707             fiz3             = _mm_macc_pd(dz32,fscal,fiz3);
 708
 709             fjx2             = _mm_macc_pd(dx32,fscal,fjx2);
 710             fjy2             = _mm_macc_pd(dy32,fscal,fjy2);
 711             fjz2             = _mm_macc_pd(dz32,fscal,fjz2);
 712
 713             /**************************
 714              * CALCULATE INTERACTIONS *
 715              **************************/
 716
 717             r33              = _mm_mul_pd(rsq33,rinv33);
 718
 719             /* EWALD ELECTROSTATICS */
 720
 721             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 722             ewrt             = _mm_mul_pd(r33,ewtabscale);
 723             ewitab           = _mm_cvttpd_epi32(ewrt);
 724 #ifdef __XOP__
 725             eweps            = _mm_frcz_pd(ewrt);
 726 #else
 727             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 728 #endif
 729             twoeweps         = _mm_add_pd(eweps,eweps);
 730             ewitab           = _mm_slli_epi32(ewitab,2);
 731             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 732             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 733             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 734             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 735             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 736             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 737             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 738             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 739             velec            = _mm_mul_pd(qq33,_mm_sub_pd(rinv33,velec));
 740             felec            = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
 741
 742             /* Update potential sum for this i atom from the interaction with this j atom. */
 743             velecsum         = _mm_add_pd(velecsum,velec);
 744
 745             fscal            = felec;
 746
 747             /* Update vectorial force */
 748             fix3             = _mm_macc_pd(dx33,fscal,fix3);
 749             fiy3             = _mm_macc_pd(dy33,fscal,fiy3);
 750             fiz3             = _mm_macc_pd(dz33,fscal,fiz3);
 751
 752             fjx3             = _mm_macc_pd(dx33,fscal,fjx3);
 753             fjy3             = _mm_macc_pd(dy33,fscal,fjy3);
 754             fjz3             = _mm_macc_pd(dz33,fscal,fjz3);
 755
 756             gmx_mm_decrement_4rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
 757
 758             /* Inner loop uses 458 flops */
 759         }
 760
 761         if(jidx<j_index_end)
 762         {
 763
 764             jnrA             = jjnr[jidx];
 765             j_coord_offsetA  = DIM*jnrA;
 766
 767             /* load j atom coordinates */
 768             gmx_mm_load_4rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 769                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 770                                               &jy2,&jz2,&jx3,&jy3,&jz3);
 771
 772             /* Calculate displacement vector */
 773             dx00             = _mm_sub_pd(ix0,jx0);
 774             dy00             = _mm_sub_pd(iy0,jy0);
 775             dz00             = _mm_sub_pd(iz0,jz0);
 776             dx11             = _mm_sub_pd(ix1,jx1);
 777             dy11             = _mm_sub_pd(iy1,jy1);
 778             dz11             = _mm_sub_pd(iz1,jz1);
 779             dx12             = _mm_sub_pd(ix1,jx2);
 780             dy12             = _mm_sub_pd(iy1,jy2);
 781             dz12             = _mm_sub_pd(iz1,jz2);
 782             dx13             = _mm_sub_pd(ix1,jx3);
 783             dy13             = _mm_sub_pd(iy1,jy3);
 784             dz13             = _mm_sub_pd(iz1,jz3);
 785             dx21             = _mm_sub_pd(ix2,jx1);
 786             dy21             = _mm_sub_pd(iy2,jy1);
 787             dz21             = _mm_sub_pd(iz2,jz1);
 788             dx22             = _mm_sub_pd(ix2,jx2);
 789             dy22             = _mm_sub_pd(iy2,jy2);
 790             dz22             = _mm_sub_pd(iz2,jz2);
 791             dx23             = _mm_sub_pd(ix2,jx3);
 792             dy23             = _mm_sub_pd(iy2,jy3);
 793             dz23             = _mm_sub_pd(iz2,jz3);
 794             dx31             = _mm_sub_pd(ix3,jx1);
 795             dy31             = _mm_sub_pd(iy3,jy1);
 796             dz31             = _mm_sub_pd(iz3,jz1);
 797             dx32             = _mm_sub_pd(ix3,jx2);
 798             dy32             = _mm_sub_pd(iy3,jy2);
 799             dz32             = _mm_sub_pd(iz3,jz2);
 800             dx33             = _mm_sub_pd(ix3,jx3);
 801             dy33             = _mm_sub_pd(iy3,jy3);
 802             dz33             = _mm_sub_pd(iz3,jz3);
 803
 804             /* Calculate squared distance and things based on it */
 805             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 806             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
 807             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
 808             rsq13            = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
 809             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
 810             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
 811             rsq23            = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
 812             rsq31            = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
 813             rsq32            = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
 814             rsq33            = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
 815
 816             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 817             rinv11           = gmx_mm_invsqrt_pd(rsq11);
 818             rinv12           = gmx_mm_invsqrt_pd(rsq12);
 819             rinv13           = gmx_mm_invsqrt_pd(rsq13);
 820             rinv21           = gmx_mm_invsqrt_pd(rsq21);
 821             rinv22           = gmx_mm_invsqrt_pd(rsq22);
 822             rinv23           = gmx_mm_invsqrt_pd(rsq23);
 823             rinv31           = gmx_mm_invsqrt_pd(rsq31);
 824             rinv32           = gmx_mm_invsqrt_pd(rsq32);
 825             rinv33           = gmx_mm_invsqrt_pd(rsq33);
 826
 827             rinvsq11         = _mm_mul_pd(rinv11,rinv11);
 828             rinvsq12         = _mm_mul_pd(rinv12,rinv12);
 829             rinvsq13         = _mm_mul_pd(rinv13,rinv13);
 830             rinvsq21         = _mm_mul_pd(rinv21,rinv21);
 831             rinvsq22         = _mm_mul_pd(rinv22,rinv22);
 832             rinvsq23         = _mm_mul_pd(rinv23,rinv23);
 833             rinvsq31         = _mm_mul_pd(rinv31,rinv31);
 834             rinvsq32         = _mm_mul_pd(rinv32,rinv32);
 835             rinvsq33         = _mm_mul_pd(rinv33,rinv33);
 836
 837             fjx0             = _mm_setzero_pd();
 838             fjy0             = _mm_setzero_pd();
 839             fjz0             = _mm_setzero_pd();
 840             fjx1             = _mm_setzero_pd();
 841             fjy1             = _mm_setzero_pd();
 842             fjz1             = _mm_setzero_pd();
 843             fjx2             = _mm_setzero_pd();
 844             fjy2             = _mm_setzero_pd();
 845             fjz2             = _mm_setzero_pd();
 846             fjx3             = _mm_setzero_pd();
 847             fjy3             = _mm_setzero_pd();
 848             fjz3             = _mm_setzero_pd();
 849
 850             /**************************
 851              * CALCULATE INTERACTIONS *
 852              **************************/
 853
 854             r00              = _mm_mul_pd(rsq00,rinv00);
 855
 856             /* Calculate table index by multiplying r with table scale and truncate to integer */
 857             rt               = _mm_mul_pd(r00,vftabscale);
 858             vfitab           = _mm_cvttpd_epi32(rt);
 859 #ifdef __XOP__
 860             vfeps            = _mm_frcz_pd(rt);
 861 #else
 862             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 863 #endif
 864             twovfeps         = _mm_add_pd(vfeps,vfeps);
 865             vfitab           = _mm_slli_epi32(vfitab,3);
 866
 867             /* CUBIC SPLINE TABLE DISPERSION */
 868             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 869             F                = _mm_setzero_pd();
 870             GMX_MM_TRANSPOSE2_PD(Y,F);
 871             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 872             H                = _mm_setzero_pd();
 873             GMX_MM_TRANSPOSE2_PD(G,H);
 874             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 875             VV               = _mm_macc_pd(vfeps,Fp,Y);
 876             vvdw6            = _mm_mul_pd(c6_00,VV);
 877             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 878             fvdw6            = _mm_mul_pd(c6_00,FF);
 879
 880             /* CUBIC SPLINE TABLE REPULSION */
 881             vfitab           = _mm_add_epi32(vfitab,ifour);
 882             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 883             F                = _mm_setzero_pd();
 884             GMX_MM_TRANSPOSE2_PD(Y,F);
 885             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 886             H                = _mm_setzero_pd();
 887             GMX_MM_TRANSPOSE2_PD(G,H);
 888             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 889             VV               = _mm_macc_pd(vfeps,Fp,Y);
 890             vvdw12           = _mm_mul_pd(c12_00,VV);
 891             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 892             fvdw12           = _mm_mul_pd(c12_00,FF);
 893             vvdw             = _mm_add_pd(vvdw12,vvdw6);
 894             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
 895
 896             /* Update potential sum for this i atom from the interaction with this j atom. */
 897             vvdw             = _mm_unpacklo_pd(vvdw,_mm_setzero_pd());
 898             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 899
 900             fscal            = fvdw;
 901
 902             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 903
 904             /* Update vectorial force */
 905             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 906             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 907             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 908
 909             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
 910             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
 911             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
 912
 913             /**************************
 914              * CALCULATE INTERACTIONS *
 915              **************************/
 916
 917             r11              = _mm_mul_pd(rsq11,rinv11);
 918
 919             /* EWALD ELECTROSTATICS */
 920
 921             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 922             ewrt             = _mm_mul_pd(r11,ewtabscale);
 923             ewitab           = _mm_cvttpd_epi32(ewrt);
 924 #ifdef __XOP__
 925             eweps            = _mm_frcz_pd(ewrt);
 926 #else
 927             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 928 #endif
 929             twoeweps         = _mm_add_pd(eweps,eweps);
 930             ewitab           = _mm_slli_epi32(ewitab,2);
 931             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 932             ewtabD           = _mm_setzero_pd();
 933             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 934             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 935             ewtabFn          = _mm_setzero_pd();
 936             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 937             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 938             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 939             velec            = _mm_mul_pd(qq11,_mm_sub_pd(rinv11,velec));
 940             felec            = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
 941
 942             /* Update potential sum for this i atom from the interaction with this j atom. */
 943             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 944             velecsum         = _mm_add_pd(velecsum,velec);
 945
 946             fscal            = felec;
 947
 948             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 949
 950             /* Update vectorial force */
 951             fix1             = _mm_macc_pd(dx11,fscal,fix1);
 952             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
 953             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
 954
 955             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
 956             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
 957             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
 958
 959             /**************************
 960              * CALCULATE INTERACTIONS *
 961              **************************/
 962
 963             r12              = _mm_mul_pd(rsq12,rinv12);
 964
 965             /* EWALD ELECTROSTATICS */
 966
 967             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 968             ewrt             = _mm_mul_pd(r12,ewtabscale);
 969             ewitab           = _mm_cvttpd_epi32(ewrt);
 970 #ifdef __XOP__
 971             eweps            = _mm_frcz_pd(ewrt);
 972 #else
 973             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 974 #endif
 975             twoeweps         = _mm_add_pd(eweps,eweps);
 976             ewitab           = _mm_slli_epi32(ewitab,2);
 977             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 978             ewtabD           = _mm_setzero_pd();
 979             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 980             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 981             ewtabFn          = _mm_setzero_pd();
 982             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 983             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 984             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 985             velec            = _mm_mul_pd(qq12,_mm_sub_pd(rinv12,velec));
 986             felec            = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
 987
 988             /* Update potential sum for this i atom from the interaction with this j atom. */
 989             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 990             velecsum         = _mm_add_pd(velecsum,velec);
 991
 992             fscal            = felec;
 993
 994             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 995
 996             /* Update vectorial force */
 997             fix1             = _mm_macc_pd(dx12,fscal,fix1);
 998             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
 999             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
1000
1001             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
1002             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
1003             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
1004
1005             /**************************
1006              * CALCULATE INTERACTIONS *
1007              **************************/
1008
1009             r13              = _mm_mul_pd(rsq13,rinv13);
1010
1011             /* EWALD ELECTROSTATICS */
1012
1013             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1014             ewrt             = _mm_mul_pd(r13,ewtabscale);
1015             ewitab           = _mm_cvttpd_epi32(ewrt);
1016 #ifdef __XOP__
1017             eweps            = _mm_frcz_pd(ewrt);
1018 #else
1019             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1020 #endif
1021             twoeweps         = _mm_add_pd(eweps,eweps);
1022             ewitab           = _mm_slli_epi32(ewitab,2);
1023             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1024             ewtabD           = _mm_setzero_pd();
1025             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1026             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1027             ewtabFn          = _mm_setzero_pd();
1028             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1029             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1030             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1031             velec            = _mm_mul_pd(qq13,_mm_sub_pd(rinv13,velec));
1032             felec            = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
1033
1034             /* Update potential sum for this i atom from the interaction with this j atom. */
1035             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1036             velecsum         = _mm_add_pd(velecsum,velec);
1037
1038             fscal            = felec;
1039
1040             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1041
1042             /* Update vectorial force */
1043             fix1             = _mm_macc_pd(dx13,fscal,fix1);
1044             fiy1             = _mm_macc_pd(dy13,fscal,fiy1);
1045             fiz1             = _mm_macc_pd(dz13,fscal,fiz1);
1046
1047             fjx3             = _mm_macc_pd(dx13,fscal,fjx3);
1048             fjy3             = _mm_macc_pd(dy13,fscal,fjy3);
1049             fjz3             = _mm_macc_pd(dz13,fscal,fjz3);
1050
1051             /**************************
1052              * CALCULATE INTERACTIONS *
1053              **************************/
1054
1055             r21              = _mm_mul_pd(rsq21,rinv21);
1056
1057             /* EWALD ELECTROSTATICS */
1058
1059             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1060             ewrt             = _mm_mul_pd(r21,ewtabscale);
1061             ewitab           = _mm_cvttpd_epi32(ewrt);
1062 #ifdef __XOP__
1063             eweps            = _mm_frcz_pd(ewrt);
1064 #else
1065             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1066 #endif
1067             twoeweps         = _mm_add_pd(eweps,eweps);
1068             ewitab           = _mm_slli_epi32(ewitab,2);
1069             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1070             ewtabD           = _mm_setzero_pd();
1071             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1072             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1073             ewtabFn          = _mm_setzero_pd();
1074             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1075             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1076             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1077             velec            = _mm_mul_pd(qq21,_mm_sub_pd(rinv21,velec));
1078             felec            = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
1079
1080             /* Update potential sum for this i atom from the interaction with this j atom. */
1081             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1082             velecsum         = _mm_add_pd(velecsum,velec);
1083
1084             fscal            = felec;
1085
1086             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1087
1088             /* Update vectorial force */
1089             fix2             = _mm_macc_pd(dx21,fscal,fix2);
1090             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
1091             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
1092
1093             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
1094             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
1095             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
1096
1097             /**************************
1098              * CALCULATE INTERACTIONS *
1099              **************************/
1100
1101             r22              = _mm_mul_pd(rsq22,rinv22);
1102
1103             /* EWALD ELECTROSTATICS */
1104
1105             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1106             ewrt             = _mm_mul_pd(r22,ewtabscale);
1107             ewitab           = _mm_cvttpd_epi32(ewrt);
1108 #ifdef __XOP__
1109             eweps            = _mm_frcz_pd(ewrt);
1110 #else
1111             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1112 #endif
1113             twoeweps         = _mm_add_pd(eweps,eweps);
1114             ewitab           = _mm_slli_epi32(ewitab,2);
1115             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1116             ewtabD           = _mm_setzero_pd();
1117             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1118             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1119             ewtabFn          = _mm_setzero_pd();
1120             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1121             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1122             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1123             velec            = _mm_mul_pd(qq22,_mm_sub_pd(rinv22,velec));
1124             felec            = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
1125
1126             /* Update potential sum for this i atom from the interaction with this j atom. */
1127             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1128             velecsum         = _mm_add_pd(velecsum,velec);
1129
1130             fscal            = felec;
1131
1132             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1133
1134             /* Update vectorial force */
1135             fix2             = _mm_macc_pd(dx22,fscal,fix2);
1136             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
1137             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
1138
1139             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
1140             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
1141             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
1142
1143             /**************************
1144              * CALCULATE INTERACTIONS *
1145              **************************/
1146
1147             r23              = _mm_mul_pd(rsq23,rinv23);
1148
1149             /* EWALD ELECTROSTATICS */
1150
1151             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1152             ewrt             = _mm_mul_pd(r23,ewtabscale);
1153             ewitab           = _mm_cvttpd_epi32(ewrt);
1154 #ifdef __XOP__
1155             eweps            = _mm_frcz_pd(ewrt);
1156 #else
1157             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1158 #endif
1159             twoeweps         = _mm_add_pd(eweps,eweps);
1160             ewitab           = _mm_slli_epi32(ewitab,2);
1161             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1162             ewtabD           = _mm_setzero_pd();
1163             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1164             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1165             ewtabFn          = _mm_setzero_pd();
1166             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1167             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1168             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1169             velec            = _mm_mul_pd(qq23,_mm_sub_pd(rinv23,velec));
1170             felec            = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
1171
1172             /* Update potential sum for this i atom from the interaction with this j atom. */
1173             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1174             velecsum         = _mm_add_pd(velecsum,velec);
1175
1176             fscal            = felec;
1177
1178             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1179
1180             /* Update vectorial force */
1181             fix2             = _mm_macc_pd(dx23,fscal,fix2);
1182             fiy2             = _mm_macc_pd(dy23,fscal,fiy2);
1183             fiz2             = _mm_macc_pd(dz23,fscal,fiz2);
1184
1185             fjx3             = _mm_macc_pd(dx23,fscal,fjx3);
1186             fjy3             = _mm_macc_pd(dy23,fscal,fjy3);
1187             fjz3             = _mm_macc_pd(dz23,fscal,fjz3);
1188
1189             /**************************
1190              * CALCULATE INTERACTIONS *
1191              **************************/
1192
1193             r31              = _mm_mul_pd(rsq31,rinv31);
1194
1195             /* EWALD ELECTROSTATICS */
1196
1197             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1198             ewrt             = _mm_mul_pd(r31,ewtabscale);
1199             ewitab           = _mm_cvttpd_epi32(ewrt);
1200 #ifdef __XOP__
1201             eweps            = _mm_frcz_pd(ewrt);
1202 #else
1203             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1204 #endif
1205             twoeweps         = _mm_add_pd(eweps,eweps);
1206             ewitab           = _mm_slli_epi32(ewitab,2);
1207             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1208             ewtabD           = _mm_setzero_pd();
1209             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1210             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1211             ewtabFn          = _mm_setzero_pd();
1212             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1213             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1214             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1215             velec            = _mm_mul_pd(qq31,_mm_sub_pd(rinv31,velec));
1216             felec            = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
1217
1218             /* Update potential sum for this i atom from the interaction with this j atom. */
1219             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1220             velecsum         = _mm_add_pd(velecsum,velec);
1221
1222             fscal            = felec;
1223
1224             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1225
1226             /* Update vectorial force */
1227             fix3             = _mm_macc_pd(dx31,fscal,fix3);
1228             fiy3             = _mm_macc_pd(dy31,fscal,fiy3);
1229             fiz3             = _mm_macc_pd(dz31,fscal,fiz3);
1230
1231             fjx1             = _mm_macc_pd(dx31,fscal,fjx1);
1232             fjy1             = _mm_macc_pd(dy31,fscal,fjy1);
1233             fjz1             = _mm_macc_pd(dz31,fscal,fjz1);
1234
1235             /**************************
1236              * CALCULATE INTERACTIONS *
1237              **************************/
1238
1239             r32              = _mm_mul_pd(rsq32,rinv32);
1240
1241             /* EWALD ELECTROSTATICS */
1242
1243             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1244             ewrt             = _mm_mul_pd(r32,ewtabscale);
1245             ewitab           = _mm_cvttpd_epi32(ewrt);
1246 #ifdef __XOP__
1247             eweps            = _mm_frcz_pd(ewrt);
1248 #else
1249             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1250 #endif
1251             twoeweps         = _mm_add_pd(eweps,eweps);
1252             ewitab           = _mm_slli_epi32(ewitab,2);
1253             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1254             ewtabD           = _mm_setzero_pd();
1255             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1256             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1257             ewtabFn          = _mm_setzero_pd();
1258             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1259             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1260             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1261             velec            = _mm_mul_pd(qq32,_mm_sub_pd(rinv32,velec));
1262             felec            = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
1263
1264             /* Update potential sum for this i atom from the interaction with this j atom. */
1265             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1266             velecsum         = _mm_add_pd(velecsum,velec);
1267
1268             fscal            = felec;
1269
1270             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1271
1272             /* Update vectorial force */
1273             fix3             = _mm_macc_pd(dx32,fscal,fix3);
1274             fiy3             = _mm_macc_pd(dy32,fscal,fiy3);
1275             fiz3             = _mm_macc_pd(dz32,fscal,fiz3);
1276
1277             fjx2             = _mm_macc_pd(dx32,fscal,fjx2);
1278             fjy2             = _mm_macc_pd(dy32,fscal,fjy2);
1279             fjz2             = _mm_macc_pd(dz32,fscal,fjz2);
1280
1281             /**************************
1282              * CALCULATE INTERACTIONS *
1283              **************************/
1284
1285             r33              = _mm_mul_pd(rsq33,rinv33);
1286
1287             /* EWALD ELECTROSTATICS */
1288
1289             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1290             ewrt             = _mm_mul_pd(r33,ewtabscale);
1291             ewitab           = _mm_cvttpd_epi32(ewrt);
1292 #ifdef __XOP__
1293             eweps            = _mm_frcz_pd(ewrt);
1294 #else
1295             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1296 #endif
1297             twoeweps         = _mm_add_pd(eweps,eweps);
1298             ewitab           = _mm_slli_epi32(ewitab,2);
1299             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1300             ewtabD           = _mm_setzero_pd();
1301             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1302             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1303             ewtabFn          = _mm_setzero_pd();
1304             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1305             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1306             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1307             velec            = _mm_mul_pd(qq33,_mm_sub_pd(rinv33,velec));
1308             felec            = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
1309
1310             /* Update potential sum for this i atom from the interaction with this j atom. */
1311             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1312             velecsum         = _mm_add_pd(velecsum,velec);
1313
1314             fscal            = felec;
1315
1316             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1317
1318             /* Update vectorial force */
1319             fix3             = _mm_macc_pd(dx33,fscal,fix3);
1320             fiy3             = _mm_macc_pd(dy33,fscal,fiy3);
1321             fiz3             = _mm_macc_pd(dz33,fscal,fiz3);
1322
1323             fjx3             = _mm_macc_pd(dx33,fscal,fjx3);
1324             fjy3             = _mm_macc_pd(dy33,fscal,fjy3);
1325             fjz3             = _mm_macc_pd(dz33,fscal,fjz3);
1326
1327             gmx_mm_decrement_4rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1328
1329             /* Inner loop uses 458 flops */
1330         }
1331
1332         /* End of innermost loop */
1333
1334         gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1335                                               f+i_coord_offset,fshift+i_shift_offset);
1336
1337         ggid                        = gid[iidx];
1338         /* Update potential energies */
1339         gmx_mm_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1340         gmx_mm_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
1341
1342         /* Increment number of inner iterations */
1343         inneriter                  += j_index_end - j_index_start;
1344
1345         /* Outer loop uses 26 flops */
1346     }
1347
1348     /* Increment number of outer iterations */
1349     outeriter        += nri;
1350
1351     /* Update outer/inner flops */
1352
1353     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*26 + inneriter*458);
1354 }
1355 /*
1356  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwCSTab_GeomW4W4_F_avx_128_fma_double
1357  * Electrostatics interaction: Ewald
1358  * VdW interaction:            CubicSplineTable
1359  * Geometry:                   Water4-Water4
1360  * Calculate force/pot:        Force
1361  */
1362 void
1363 nb_kernel_ElecEw_VdwCSTab_GeomW4W4_F_avx_128_fma_double
1364                     (t_nblist                    * gmx_restrict       nlist,
1365                      rvec                        * gmx_restrict          xx,
1366                      rvec                        * gmx_restrict          ff,
1367                      t_forcerec                  * gmx_restrict          fr,
1368                      t_mdatoms                   * gmx_restrict     mdatoms,
1369                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1370                      t_nrnb                      * gmx_restrict        nrnb)
1371 {
1372     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1373      * just 0 for non-waters.
1374      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
1375      * jnr indices corresponding to data put in the four positions in the SIMD register.
1376      */
1377     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
1378     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1379     int              jnrA,jnrB;
1380     int              j_coord_offsetA,j_coord_offsetB;
1381     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
1382     real             rcutoff_scalar;
1383     real             *shiftvec,*fshift,*x,*f;
1384     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1385     int              vdwioffset0;
1386     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1387     int              vdwioffset1;
1388     __m128d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1389     int              vdwioffset2;
1390     __m128d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1391     int              vdwioffset3;
1392     __m128d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
1393     int              vdwjidx0A,vdwjidx0B;
1394     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1395     int              vdwjidx1A,vdwjidx1B;
1396     __m128d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1397     int              vdwjidx2A,vdwjidx2B;
1398     __m128d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1399     int              vdwjidx3A,vdwjidx3B;
1400     __m128d          jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
1401     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1402     __m128d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1403     __m128d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1404     __m128d          dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
1405     __m128d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1406     __m128d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1407     __m128d          dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
1408     __m128d          dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
1409     __m128d          dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
1410     __m128d          dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
1411     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
1412     real             *charge;
1413     int              nvdwtype;
1414     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1415     int              *vdwtype;
1416     real             *vdwparam;
1417     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
1418     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
1419     __m128i          vfitab;
1420     __m128i          ifour       = _mm_set1_epi32(4);
1421     __m128d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF,twovfeps;
1422     real             *vftab;
1423     __m128i          ewitab;
1424     __m128d          ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
1425     real             *ewtab;
1426     __m128d          dummy_mask,cutoff_mask;
1427     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
1428     __m128d          one     = _mm_set1_pd(1.0);
1429     __m128d          two     = _mm_set1_pd(2.0);
1430     x                = xx[0];
1431     f                = ff[0];
1432
1433     nri              = nlist->nri;
1434     iinr             = nlist->iinr;
1435     jindex           = nlist->jindex;
1436     jjnr             = nlist->jjnr;
1437     shiftidx         = nlist->shift;
1438     gid              = nlist->gid;
1439     shiftvec         = fr->shift_vec[0];
1440     fshift           = fr->fshift[0];
1441     facel            = _mm_set1_pd(fr->epsfac);
1442     charge           = mdatoms->chargeA;
1443     nvdwtype         = fr->ntype;
1444     vdwparam         = fr->nbfp;
1445     vdwtype          = mdatoms->typeA;
1446
1447     vftab            = kernel_data->table_vdw->data;
1448     vftabscale       = _mm_set1_pd(kernel_data->table_vdw->scale);
1449
1450     sh_ewald         = _mm_set1_pd(fr->ic->sh_ewald);
1451     ewtab            = fr->ic->tabq_coul_F;
1452     ewtabscale       = _mm_set1_pd(fr->ic->tabq_scale);
1453     ewtabhalfspace   = _mm_set1_pd(0.5/fr->ic->tabq_scale);
1454
1455     /* Setup water-specific parameters */
1456     inr              = nlist->iinr[0];
1457     iq1              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
1458     iq2              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
1459     iq3              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
1460     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
1461
1462     jq1              = _mm_set1_pd(charge[inr+1]);
1463     jq2              = _mm_set1_pd(charge[inr+2]);
1464     jq3              = _mm_set1_pd(charge[inr+3]);
1465     vdwjidx0A        = 2*vdwtype[inr+0];
1466     c6_00            = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A]);
1467     c12_00           = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A+1]);
1468     qq11             = _mm_mul_pd(iq1,jq1);
1469     qq12             = _mm_mul_pd(iq1,jq2);
1470     qq13             = _mm_mul_pd(iq1,jq3);
1471     qq21             = _mm_mul_pd(iq2,jq1);
1472     qq22             = _mm_mul_pd(iq2,jq2);
1473     qq23             = _mm_mul_pd(iq2,jq3);
1474     qq31             = _mm_mul_pd(iq3,jq1);
1475     qq32             = _mm_mul_pd(iq3,jq2);
1476     qq33             = _mm_mul_pd(iq3,jq3);
1477
1478     /* Avoid stupid compiler warnings */
1479     jnrA = jnrB = 0;
1480     j_coord_offsetA = 0;
1481     j_coord_offsetB = 0;
1482
1483     outeriter        = 0;
1484     inneriter        = 0;
1485
1486     /* Start outer loop over neighborlists */
1487     for(iidx=0; iidx<nri; iidx++)
1488     {
1489         /* Load shift vector for this list */
1490         i_shift_offset   = DIM*shiftidx[iidx];
1491
1492         /* Load limits for loop over neighbors */
1493         j_index_start    = jindex[iidx];
1494         j_index_end      = jindex[iidx+1];
1495
1496         /* Get outer coordinate index */
1497         inr              = iinr[iidx];
1498         i_coord_offset   = DIM*inr;
1499
1500         /* Load i particle coords and add shift vector */
1501         gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1502                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1503
1504         fix0             = _mm_setzero_pd();
1505         fiy0             = _mm_setzero_pd();
1506         fiz0             = _mm_setzero_pd();
1507         fix1             = _mm_setzero_pd();
1508         fiy1             = _mm_setzero_pd();
1509         fiz1             = _mm_setzero_pd();
1510         fix2             = _mm_setzero_pd();
1511         fiy2             = _mm_setzero_pd();
1512         fiz2             = _mm_setzero_pd();
1513         fix3             = _mm_setzero_pd();
1514         fiy3             = _mm_setzero_pd();
1515         fiz3             = _mm_setzero_pd();
1516
1517         /* Start inner kernel loop */
1518         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
1519         {
1520
1521             /* Get j neighbor index, and coordinate index */
1522             jnrA             = jjnr[jidx];
1523             jnrB             = jjnr[jidx+1];
1524             j_coord_offsetA  = DIM*jnrA;
1525             j_coord_offsetB  = DIM*jnrB;
1526
1527             /* load j atom coordinates */
1528             gmx_mm_load_4rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1529                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1530                                               &jy2,&jz2,&jx3,&jy3,&jz3);
1531
1532             /* Calculate displacement vector */
1533             dx00             = _mm_sub_pd(ix0,jx0);
1534             dy00             = _mm_sub_pd(iy0,jy0);
1535             dz00             = _mm_sub_pd(iz0,jz0);
1536             dx11             = _mm_sub_pd(ix1,jx1);
1537             dy11             = _mm_sub_pd(iy1,jy1);
1538             dz11             = _mm_sub_pd(iz1,jz1);
1539             dx12             = _mm_sub_pd(ix1,jx2);
1540             dy12             = _mm_sub_pd(iy1,jy2);
1541             dz12             = _mm_sub_pd(iz1,jz2);
1542             dx13             = _mm_sub_pd(ix1,jx3);
1543             dy13             = _mm_sub_pd(iy1,jy3);
1544             dz13             = _mm_sub_pd(iz1,jz3);
1545             dx21             = _mm_sub_pd(ix2,jx1);
1546             dy21             = _mm_sub_pd(iy2,jy1);
1547             dz21             = _mm_sub_pd(iz2,jz1);
1548             dx22             = _mm_sub_pd(ix2,jx2);
1549             dy22             = _mm_sub_pd(iy2,jy2);
1550             dz22             = _mm_sub_pd(iz2,jz2);
1551             dx23             = _mm_sub_pd(ix2,jx3);
1552             dy23             = _mm_sub_pd(iy2,jy3);
1553             dz23             = _mm_sub_pd(iz2,jz3);
1554             dx31             = _mm_sub_pd(ix3,jx1);
1555             dy31             = _mm_sub_pd(iy3,jy1);
1556             dz31             = _mm_sub_pd(iz3,jz1);
1557             dx32             = _mm_sub_pd(ix3,jx2);
1558             dy32             = _mm_sub_pd(iy3,jy2);
1559             dz32             = _mm_sub_pd(iz3,jz2);
1560             dx33             = _mm_sub_pd(ix3,jx3);
1561             dy33             = _mm_sub_pd(iy3,jy3);
1562             dz33             = _mm_sub_pd(iz3,jz3);
1563
1564             /* Calculate squared distance and things based on it */
1565             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1566             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
1567             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
1568             rsq13            = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
1569             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
1570             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
1571             rsq23            = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
1572             rsq31            = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
1573             rsq32            = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
1574             rsq33            = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
1575
1576             rinv00           = gmx_mm_invsqrt_pd(rsq00);
1577             rinv11           = gmx_mm_invsqrt_pd(rsq11);
1578             rinv12           = gmx_mm_invsqrt_pd(rsq12);
1579             rinv13           = gmx_mm_invsqrt_pd(rsq13);
1580             rinv21           = gmx_mm_invsqrt_pd(rsq21);
1581             rinv22           = gmx_mm_invsqrt_pd(rsq22);
1582             rinv23           = gmx_mm_invsqrt_pd(rsq23);
1583             rinv31           = gmx_mm_invsqrt_pd(rsq31);
1584             rinv32           = gmx_mm_invsqrt_pd(rsq32);
1585             rinv33           = gmx_mm_invsqrt_pd(rsq33);
1586
1587             rinvsq11         = _mm_mul_pd(rinv11,rinv11);
1588             rinvsq12         = _mm_mul_pd(rinv12,rinv12);
1589             rinvsq13         = _mm_mul_pd(rinv13,rinv13);
1590             rinvsq21         = _mm_mul_pd(rinv21,rinv21);
1591             rinvsq22         = _mm_mul_pd(rinv22,rinv22);
1592             rinvsq23         = _mm_mul_pd(rinv23,rinv23);
1593             rinvsq31         = _mm_mul_pd(rinv31,rinv31);
1594             rinvsq32         = _mm_mul_pd(rinv32,rinv32);
1595             rinvsq33         = _mm_mul_pd(rinv33,rinv33);
1596
1597             fjx0             = _mm_setzero_pd();
1598             fjy0             = _mm_setzero_pd();
1599             fjz0             = _mm_setzero_pd();
1600             fjx1             = _mm_setzero_pd();
1601             fjy1             = _mm_setzero_pd();
1602             fjz1             = _mm_setzero_pd();
1603             fjx2             = _mm_setzero_pd();
1604             fjy2             = _mm_setzero_pd();
1605             fjz2             = _mm_setzero_pd();
1606             fjx3             = _mm_setzero_pd();
1607             fjy3             = _mm_setzero_pd();
1608             fjz3             = _mm_setzero_pd();
1609
1610             /**************************
1611              * CALCULATE INTERACTIONS *
1612              **************************/
1613
1614             r00              = _mm_mul_pd(rsq00,rinv00);
1615
1616             /* Calculate table index by multiplying r with table scale and truncate to integer */
1617             rt               = _mm_mul_pd(r00,vftabscale);
1618             vfitab           = _mm_cvttpd_epi32(rt);
1619 #ifdef __XOP__
1620             vfeps            = _mm_frcz_pd(rt);
1621 #else
1622             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1623 #endif
1624             twovfeps         = _mm_add_pd(vfeps,vfeps);
1625             vfitab           = _mm_slli_epi32(vfitab,3);
1626
1627             /* CUBIC SPLINE TABLE DISPERSION */
1628             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1629             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1630             GMX_MM_TRANSPOSE2_PD(Y,F);
1631             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1632             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1633             GMX_MM_TRANSPOSE2_PD(G,H);
1634             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
1635             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
1636             fvdw6            = _mm_mul_pd(c6_00,FF);
1637
1638             /* CUBIC SPLINE TABLE REPULSION */
1639             vfitab           = _mm_add_epi32(vfitab,ifour);
1640             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1641             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1642             GMX_MM_TRANSPOSE2_PD(Y,F);
1643             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
1644             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
1645             GMX_MM_TRANSPOSE2_PD(G,H);
1646             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
1647             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
1648             fvdw12           = _mm_mul_pd(c12_00,FF);
1649             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
1650
1651             fscal            = fvdw;
1652
1653             /* Update vectorial force */
1654             fix0             = _mm_macc_pd(dx00,fscal,fix0);
1655             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
1656             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
1657
1658             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
1659             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
1660             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
1661
1662             /**************************
1663              * CALCULATE INTERACTIONS *
1664              **************************/
1665
1666             r11              = _mm_mul_pd(rsq11,rinv11);
1667
1668             /* EWALD ELECTROSTATICS */
1669
1670             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1671             ewrt             = _mm_mul_pd(r11,ewtabscale);
1672             ewitab           = _mm_cvttpd_epi32(ewrt);
1673 #ifdef __XOP__
1674             eweps            = _mm_frcz_pd(ewrt);
1675 #else
1676             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1677 #endif
1678             twoeweps         = _mm_add_pd(eweps,eweps);
1679             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1680                                          &ewtabF,&ewtabFn);
1681             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1682             felec            = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
1683
1684             fscal            = felec;
1685
1686             /* Update vectorial force */
1687             fix1             = _mm_macc_pd(dx11,fscal,fix1);
1688             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
1689             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
1690
1691             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
1692             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
1693             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
1694
1695             /**************************
1696              * CALCULATE INTERACTIONS *
1697              **************************/
1698
1699             r12              = _mm_mul_pd(rsq12,rinv12);
1700
1701             /* EWALD ELECTROSTATICS */
1702
1703             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1704             ewrt             = _mm_mul_pd(r12,ewtabscale);
1705             ewitab           = _mm_cvttpd_epi32(ewrt);
1706 #ifdef __XOP__
1707             eweps            = _mm_frcz_pd(ewrt);
1708 #else
1709             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1710 #endif
1711             twoeweps         = _mm_add_pd(eweps,eweps);
1712             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1713                                          &ewtabF,&ewtabFn);
1714             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1715             felec            = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
1716
1717             fscal            = felec;
1718
1719             /* Update vectorial force */
1720             fix1             = _mm_macc_pd(dx12,fscal,fix1);
1721             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
1722             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
1723
1724             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
1725             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
1726             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
1727
1728             /**************************
1729              * CALCULATE INTERACTIONS *
1730              **************************/
1731
1732             r13              = _mm_mul_pd(rsq13,rinv13);
1733
1734             /* EWALD ELECTROSTATICS */
1735
1736             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1737             ewrt             = _mm_mul_pd(r13,ewtabscale);
1738             ewitab           = _mm_cvttpd_epi32(ewrt);
1739 #ifdef __XOP__
1740             eweps            = _mm_frcz_pd(ewrt);
1741 #else
1742             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1743 #endif
1744             twoeweps         = _mm_add_pd(eweps,eweps);
1745             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1746                                          &ewtabF,&ewtabFn);
1747             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1748             felec            = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
1749
1750             fscal            = felec;
1751
1752             /* Update vectorial force */
1753             fix1             = _mm_macc_pd(dx13,fscal,fix1);
1754             fiy1             = _mm_macc_pd(dy13,fscal,fiy1);
1755             fiz1             = _mm_macc_pd(dz13,fscal,fiz1);
1756
1757             fjx3             = _mm_macc_pd(dx13,fscal,fjx3);
1758             fjy3             = _mm_macc_pd(dy13,fscal,fjy3);
1759             fjz3             = _mm_macc_pd(dz13,fscal,fjz3);
1760
1761             /**************************
1762              * CALCULATE INTERACTIONS *
1763              **************************/
1764
1765             r21              = _mm_mul_pd(rsq21,rinv21);
1766
1767             /* EWALD ELECTROSTATICS */
1768
1769             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1770             ewrt             = _mm_mul_pd(r21,ewtabscale);
1771             ewitab           = _mm_cvttpd_epi32(ewrt);
1772 #ifdef __XOP__
1773             eweps            = _mm_frcz_pd(ewrt);
1774 #else
1775             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1776 #endif
1777             twoeweps         = _mm_add_pd(eweps,eweps);
1778             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1779                                          &ewtabF,&ewtabFn);
1780             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1781             felec            = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
1782
1783             fscal            = felec;
1784
1785             /* Update vectorial force */
1786             fix2             = _mm_macc_pd(dx21,fscal,fix2);
1787             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
1788             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
1789
1790             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
1791             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
1792             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
1793
1794             /**************************
1795              * CALCULATE INTERACTIONS *
1796              **************************/
1797
1798             r22              = _mm_mul_pd(rsq22,rinv22);
1799
1800             /* EWALD ELECTROSTATICS */
1801
1802             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1803             ewrt             = _mm_mul_pd(r22,ewtabscale);
1804             ewitab           = _mm_cvttpd_epi32(ewrt);
1805 #ifdef __XOP__
1806             eweps            = _mm_frcz_pd(ewrt);
1807 #else
1808             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1809 #endif
1810             twoeweps         = _mm_add_pd(eweps,eweps);
1811             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1812                                          &ewtabF,&ewtabFn);
1813             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1814             felec            = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
1815
1816             fscal            = felec;
1817
1818             /* Update vectorial force */
1819             fix2             = _mm_macc_pd(dx22,fscal,fix2);
1820             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
1821             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
1822
1823             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
1824             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
1825             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
1826
1827             /**************************
1828              * CALCULATE INTERACTIONS *
1829              **************************/
1830
1831             r23              = _mm_mul_pd(rsq23,rinv23);
1832
1833             /* EWALD ELECTROSTATICS */
1834
1835             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1836             ewrt             = _mm_mul_pd(r23,ewtabscale);
1837             ewitab           = _mm_cvttpd_epi32(ewrt);
1838 #ifdef __XOP__
1839             eweps            = _mm_frcz_pd(ewrt);
1840 #else
1841             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1842 #endif
1843             twoeweps         = _mm_add_pd(eweps,eweps);
1844             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1845                                          &ewtabF,&ewtabFn);
1846             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1847             felec            = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
1848
1849             fscal            = felec;
1850
1851             /* Update vectorial force */
1852             fix2             = _mm_macc_pd(dx23,fscal,fix2);
1853             fiy2             = _mm_macc_pd(dy23,fscal,fiy2);
1854             fiz2             = _mm_macc_pd(dz23,fscal,fiz2);
1855
1856             fjx3             = _mm_macc_pd(dx23,fscal,fjx3);
1857             fjy3             = _mm_macc_pd(dy23,fscal,fjy3);
1858             fjz3             = _mm_macc_pd(dz23,fscal,fjz3);
1859
1860             /**************************
1861              * CALCULATE INTERACTIONS *
1862              **************************/
1863
1864             r31              = _mm_mul_pd(rsq31,rinv31);
1865
1866             /* EWALD ELECTROSTATICS */
1867
1868             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1869             ewrt             = _mm_mul_pd(r31,ewtabscale);
1870             ewitab           = _mm_cvttpd_epi32(ewrt);
1871 #ifdef __XOP__
1872             eweps            = _mm_frcz_pd(ewrt);
1873 #else
1874             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1875 #endif
1876             twoeweps         = _mm_add_pd(eweps,eweps);
1877             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1878                                          &ewtabF,&ewtabFn);
1879             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1880             felec            = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
1881
1882             fscal            = felec;
1883
1884             /* Update vectorial force */
1885             fix3             = _mm_macc_pd(dx31,fscal,fix3);
1886             fiy3             = _mm_macc_pd(dy31,fscal,fiy3);
1887             fiz3             = _mm_macc_pd(dz31,fscal,fiz3);
1888
1889             fjx1             = _mm_macc_pd(dx31,fscal,fjx1);
1890             fjy1             = _mm_macc_pd(dy31,fscal,fjy1);
1891             fjz1             = _mm_macc_pd(dz31,fscal,fjz1);
1892
1893             /**************************
1894              * CALCULATE INTERACTIONS *
1895              **************************/
1896
1897             r32              = _mm_mul_pd(rsq32,rinv32);
1898
1899             /* EWALD ELECTROSTATICS */
1900
1901             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1902             ewrt             = _mm_mul_pd(r32,ewtabscale);
1903             ewitab           = _mm_cvttpd_epi32(ewrt);
1904 #ifdef __XOP__
1905             eweps            = _mm_frcz_pd(ewrt);
1906 #else
1907             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1908 #endif
1909             twoeweps         = _mm_add_pd(eweps,eweps);
1910             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1911                                          &ewtabF,&ewtabFn);
1912             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1913             felec            = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
1914
1915             fscal            = felec;
1916
1917             /* Update vectorial force */
1918             fix3             = _mm_macc_pd(dx32,fscal,fix3);
1919             fiy3             = _mm_macc_pd(dy32,fscal,fiy3);
1920             fiz3             = _mm_macc_pd(dz32,fscal,fiz3);
1921
1922             fjx2             = _mm_macc_pd(dx32,fscal,fjx2);
1923             fjy2             = _mm_macc_pd(dy32,fscal,fjy2);
1924             fjz2             = _mm_macc_pd(dz32,fscal,fjz2);
1925
1926             /**************************
1927              * CALCULATE INTERACTIONS *
1928              **************************/
1929
1930             r33              = _mm_mul_pd(rsq33,rinv33);
1931
1932             /* EWALD ELECTROSTATICS */
1933
1934             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1935             ewrt             = _mm_mul_pd(r33,ewtabscale);
1936             ewitab           = _mm_cvttpd_epi32(ewrt);
1937 #ifdef __XOP__
1938             eweps            = _mm_frcz_pd(ewrt);
1939 #else
1940             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1941 #endif
1942             twoeweps         = _mm_add_pd(eweps,eweps);
1943             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1944                                          &ewtabF,&ewtabFn);
1945             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1946             felec            = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
1947
1948             fscal            = felec;
1949
1950             /* Update vectorial force */
1951             fix3             = _mm_macc_pd(dx33,fscal,fix3);
1952             fiy3             = _mm_macc_pd(dy33,fscal,fiy3);
1953             fiz3             = _mm_macc_pd(dz33,fscal,fiz3);
1954
1955             fjx3             = _mm_macc_pd(dx33,fscal,fjx3);
1956             fjy3             = _mm_macc_pd(dy33,fscal,fjy3);
1957             fjz3             = _mm_macc_pd(dz33,fscal,fjz3);
1958
1959             gmx_mm_decrement_4rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1960
1961             /* Inner loop uses 405 flops */
1962         }
1963
1964         if(jidx<j_index_end)
1965         {
1966
1967             jnrA             = jjnr[jidx];
1968             j_coord_offsetA  = DIM*jnrA;
1969
1970             /* load j atom coordinates */
1971             gmx_mm_load_4rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
1972                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1973                                               &jy2,&jz2,&jx3,&jy3,&jz3);
1974
1975             /* Calculate displacement vector */
1976             dx00             = _mm_sub_pd(ix0,jx0);
1977             dy00             = _mm_sub_pd(iy0,jy0);
1978             dz00             = _mm_sub_pd(iz0,jz0);
1979             dx11             = _mm_sub_pd(ix1,jx1);
1980             dy11             = _mm_sub_pd(iy1,jy1);
1981             dz11             = _mm_sub_pd(iz1,jz1);
1982             dx12             = _mm_sub_pd(ix1,jx2);
1983             dy12             = _mm_sub_pd(iy1,jy2);
1984             dz12             = _mm_sub_pd(iz1,jz2);
1985             dx13             = _mm_sub_pd(ix1,jx3);
1986             dy13             = _mm_sub_pd(iy1,jy3);
1987             dz13             = _mm_sub_pd(iz1,jz3);
1988             dx21             = _mm_sub_pd(ix2,jx1);
1989             dy21             = _mm_sub_pd(iy2,jy1);
1990             dz21             = _mm_sub_pd(iz2,jz1);
1991             dx22             = _mm_sub_pd(ix2,jx2);
1992             dy22             = _mm_sub_pd(iy2,jy2);
1993             dz22             = _mm_sub_pd(iz2,jz2);
1994             dx23             = _mm_sub_pd(ix2,jx3);
1995             dy23             = _mm_sub_pd(iy2,jy3);
1996             dz23             = _mm_sub_pd(iz2,jz3);
1997             dx31             = _mm_sub_pd(ix3,jx1);
1998             dy31             = _mm_sub_pd(iy3,jy1);
1999             dz31             = _mm_sub_pd(iz3,jz1);
2000             dx32             = _mm_sub_pd(ix3,jx2);
2001             dy32             = _mm_sub_pd(iy3,jy2);
2002             dz32             = _mm_sub_pd(iz3,jz2);
2003             dx33             = _mm_sub_pd(ix3,jx3);
2004             dy33             = _mm_sub_pd(iy3,jy3);
2005             dz33             = _mm_sub_pd(iz3,jz3);
2006
2007             /* Calculate squared distance and things based on it */
2008             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
2009             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
2010             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
2011             rsq13            = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
2012             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
2013             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
2014             rsq23            = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
2015             rsq31            = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
2016             rsq32            = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
2017             rsq33            = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
2018
2019             rinv00           = gmx_mm_invsqrt_pd(rsq00);
2020             rinv11           = gmx_mm_invsqrt_pd(rsq11);
2021             rinv12           = gmx_mm_invsqrt_pd(rsq12);
2022             rinv13           = gmx_mm_invsqrt_pd(rsq13);
2023             rinv21           = gmx_mm_invsqrt_pd(rsq21);
2024             rinv22           = gmx_mm_invsqrt_pd(rsq22);
2025             rinv23           = gmx_mm_invsqrt_pd(rsq23);
2026             rinv31           = gmx_mm_invsqrt_pd(rsq31);
2027             rinv32           = gmx_mm_invsqrt_pd(rsq32);
2028             rinv33           = gmx_mm_invsqrt_pd(rsq33);
2029
2030             rinvsq11         = _mm_mul_pd(rinv11,rinv11);
2031             rinvsq12         = _mm_mul_pd(rinv12,rinv12);
2032             rinvsq13         = _mm_mul_pd(rinv13,rinv13);
2033             rinvsq21         = _mm_mul_pd(rinv21,rinv21);
2034             rinvsq22         = _mm_mul_pd(rinv22,rinv22);
2035             rinvsq23         = _mm_mul_pd(rinv23,rinv23);
2036             rinvsq31         = _mm_mul_pd(rinv31,rinv31);
2037             rinvsq32         = _mm_mul_pd(rinv32,rinv32);
2038             rinvsq33         = _mm_mul_pd(rinv33,rinv33);
2039
2040             fjx0             = _mm_setzero_pd();
2041             fjy0             = _mm_setzero_pd();
2042             fjz0             = _mm_setzero_pd();
2043             fjx1             = _mm_setzero_pd();
2044             fjy1             = _mm_setzero_pd();
2045             fjz1             = _mm_setzero_pd();
2046             fjx2             = _mm_setzero_pd();
2047             fjy2             = _mm_setzero_pd();
2048             fjz2             = _mm_setzero_pd();
2049             fjx3             = _mm_setzero_pd();
2050             fjy3             = _mm_setzero_pd();
2051             fjz3             = _mm_setzero_pd();
2052
2053             /**************************
2054              * CALCULATE INTERACTIONS *
2055              **************************/
2056
2057             r00              = _mm_mul_pd(rsq00,rinv00);
2058
2059             /* Calculate table index by multiplying r with table scale and truncate to integer */
2060             rt               = _mm_mul_pd(r00,vftabscale);
2061             vfitab           = _mm_cvttpd_epi32(rt);
2062 #ifdef __XOP__
2063             vfeps            = _mm_frcz_pd(rt);
2064 #else
2065             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
2066 #endif
2067             twovfeps         = _mm_add_pd(vfeps,vfeps);
2068             vfitab           = _mm_slli_epi32(vfitab,3);
2069
2070             /* CUBIC SPLINE TABLE DISPERSION */
2071             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
2072             F                = _mm_setzero_pd();
2073             GMX_MM_TRANSPOSE2_PD(Y,F);
2074             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
2075             H                = _mm_setzero_pd();
2076             GMX_MM_TRANSPOSE2_PD(G,H);
2077             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
2078             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
2079             fvdw6            = _mm_mul_pd(c6_00,FF);
2080
2081             /* CUBIC SPLINE TABLE REPULSION */
2082             vfitab           = _mm_add_epi32(vfitab,ifour);
2083             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
2084             F                = _mm_setzero_pd();
2085             GMX_MM_TRANSPOSE2_PD(Y,F);
2086             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
2087             H                = _mm_setzero_pd();
2088             GMX_MM_TRANSPOSE2_PD(G,H);
2089             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
2090             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
2091             fvdw12           = _mm_mul_pd(c12_00,FF);
2092             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
2093
2094             fscal            = fvdw;
2095
2096             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2097
2098             /* Update vectorial force */
2099             fix0             = _mm_macc_pd(dx00,fscal,fix0);
2100             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
2101             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
2102
2103             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
2104             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
2105             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
2106
2107             /**************************
2108              * CALCULATE INTERACTIONS *
2109              **************************/
2110
2111             r11              = _mm_mul_pd(rsq11,rinv11);
2112
2113             /* EWALD ELECTROSTATICS */
2114
2115             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2116             ewrt             = _mm_mul_pd(r11,ewtabscale);
2117             ewitab           = _mm_cvttpd_epi32(ewrt);
2118 #ifdef __XOP__
2119             eweps            = _mm_frcz_pd(ewrt);
2120 #else
2121             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2122 #endif
2123             twoeweps         = _mm_add_pd(eweps,eweps);
2124             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2125             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2126             felec            = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
2127
2128             fscal            = felec;
2129
2130             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2131
2132             /* Update vectorial force */
2133             fix1             = _mm_macc_pd(dx11,fscal,fix1);
2134             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
2135             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
2136
2137             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
2138             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
2139             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
2140
2141             /**************************
2142              * CALCULATE INTERACTIONS *
2143              **************************/
2144
2145             r12              = _mm_mul_pd(rsq12,rinv12);
2146
2147             /* EWALD ELECTROSTATICS */
2148
2149             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2150             ewrt             = _mm_mul_pd(r12,ewtabscale);
2151             ewitab           = _mm_cvttpd_epi32(ewrt);
2152 #ifdef __XOP__
2153             eweps            = _mm_frcz_pd(ewrt);
2154 #else
2155             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2156 #endif
2157             twoeweps         = _mm_add_pd(eweps,eweps);
2158             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2159             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2160             felec            = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
2161
2162             fscal            = felec;
2163
2164             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2165
2166             /* Update vectorial force */
2167             fix1             = _mm_macc_pd(dx12,fscal,fix1);
2168             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
2169             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
2170
2171             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
2172             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
2173             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
2174
2175             /**************************
2176              * CALCULATE INTERACTIONS *
2177              **************************/
2178
2179             r13              = _mm_mul_pd(rsq13,rinv13);
2180
2181             /* EWALD ELECTROSTATICS */
2182
2183             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2184             ewrt             = _mm_mul_pd(r13,ewtabscale);
2185             ewitab           = _mm_cvttpd_epi32(ewrt);
2186 #ifdef __XOP__
2187             eweps            = _mm_frcz_pd(ewrt);
2188 #else
2189             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2190 #endif
2191             twoeweps         = _mm_add_pd(eweps,eweps);
2192             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2193             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2194             felec            = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
2195
2196             fscal            = felec;
2197
2198             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2199
2200             /* Update vectorial force */
2201             fix1             = _mm_macc_pd(dx13,fscal,fix1);
2202             fiy1             = _mm_macc_pd(dy13,fscal,fiy1);
2203             fiz1             = _mm_macc_pd(dz13,fscal,fiz1);
2204
2205             fjx3             = _mm_macc_pd(dx13,fscal,fjx3);
2206             fjy3             = _mm_macc_pd(dy13,fscal,fjy3);
2207             fjz3             = _mm_macc_pd(dz13,fscal,fjz3);
2208
2209             /**************************
2210              * CALCULATE INTERACTIONS *
2211              **************************/
2212
2213             r21              = _mm_mul_pd(rsq21,rinv21);
2214
2215             /* EWALD ELECTROSTATICS */
2216
2217             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2218             ewrt             = _mm_mul_pd(r21,ewtabscale);
2219             ewitab           = _mm_cvttpd_epi32(ewrt);
2220 #ifdef __XOP__
2221             eweps            = _mm_frcz_pd(ewrt);
2222 #else
2223             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2224 #endif
2225             twoeweps         = _mm_add_pd(eweps,eweps);
2226             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2227             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2228             felec            = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
2229
2230             fscal            = felec;
2231
2232             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2233
2234             /* Update vectorial force */
2235             fix2             = _mm_macc_pd(dx21,fscal,fix2);
2236             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
2237             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
2238
2239             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
2240             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
2241             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
2242
2243             /**************************
2244              * CALCULATE INTERACTIONS *
2245              **************************/
2246
2247             r22              = _mm_mul_pd(rsq22,rinv22);
2248
2249             /* EWALD ELECTROSTATICS */
2250
2251             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2252             ewrt             = _mm_mul_pd(r22,ewtabscale);
2253             ewitab           = _mm_cvttpd_epi32(ewrt);
2254 #ifdef __XOP__
2255             eweps            = _mm_frcz_pd(ewrt);
2256 #else
2257             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2258 #endif
2259             twoeweps         = _mm_add_pd(eweps,eweps);
2260             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2261             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2262             felec            = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
2263
2264             fscal            = felec;
2265
2266             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2267
2268             /* Update vectorial force */
2269             fix2             = _mm_macc_pd(dx22,fscal,fix2);
2270             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
2271             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
2272
2273             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
2274             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
2275             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
2276
2277             /**************************
2278              * CALCULATE INTERACTIONS *
2279              **************************/
2280
2281             r23              = _mm_mul_pd(rsq23,rinv23);
2282
2283             /* EWALD ELECTROSTATICS */
2284
2285             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2286             ewrt             = _mm_mul_pd(r23,ewtabscale);
2287             ewitab           = _mm_cvttpd_epi32(ewrt);
2288 #ifdef __XOP__
2289             eweps            = _mm_frcz_pd(ewrt);
2290 #else
2291             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2292 #endif
2293             twoeweps         = _mm_add_pd(eweps,eweps);
2294             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2295             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2296             felec            = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
2297
2298             fscal            = felec;
2299
2300             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2301
2302             /* Update vectorial force */
2303             fix2             = _mm_macc_pd(dx23,fscal,fix2);
2304             fiy2             = _mm_macc_pd(dy23,fscal,fiy2);
2305             fiz2             = _mm_macc_pd(dz23,fscal,fiz2);
2306
2307             fjx3             = _mm_macc_pd(dx23,fscal,fjx3);
2308             fjy3             = _mm_macc_pd(dy23,fscal,fjy3);
2309             fjz3             = _mm_macc_pd(dz23,fscal,fjz3);
2310
2311             /**************************
2312              * CALCULATE INTERACTIONS *
2313              **************************/
2314
2315             r31              = _mm_mul_pd(rsq31,rinv31);
2316
2317             /* EWALD ELECTROSTATICS */
2318
2319             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2320             ewrt             = _mm_mul_pd(r31,ewtabscale);
2321             ewitab           = _mm_cvttpd_epi32(ewrt);
2322 #ifdef __XOP__
2323             eweps            = _mm_frcz_pd(ewrt);
2324 #else
2325             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2326 #endif
2327             twoeweps         = _mm_add_pd(eweps,eweps);
2328             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2329             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2330             felec            = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
2331
2332             fscal            = felec;
2333
2334             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2335
2336             /* Update vectorial force */
2337             fix3             = _mm_macc_pd(dx31,fscal,fix3);
2338             fiy3             = _mm_macc_pd(dy31,fscal,fiy3);
2339             fiz3             = _mm_macc_pd(dz31,fscal,fiz3);
2340
2341             fjx1             = _mm_macc_pd(dx31,fscal,fjx1);
2342             fjy1             = _mm_macc_pd(dy31,fscal,fjy1);
2343             fjz1             = _mm_macc_pd(dz31,fscal,fjz1);
2344
2345             /**************************
2346              * CALCULATE INTERACTIONS *
2347              **************************/
2348
2349             r32              = _mm_mul_pd(rsq32,rinv32);
2350
2351             /* EWALD ELECTROSTATICS */
2352
2353             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2354             ewrt             = _mm_mul_pd(r32,ewtabscale);
2355             ewitab           = _mm_cvttpd_epi32(ewrt);
2356 #ifdef __XOP__
2357             eweps            = _mm_frcz_pd(ewrt);
2358 #else
2359             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2360 #endif
2361             twoeweps         = _mm_add_pd(eweps,eweps);
2362             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2363             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2364             felec            = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
2365
2366             fscal            = felec;
2367
2368             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2369
2370             /* Update vectorial force */
2371             fix3             = _mm_macc_pd(dx32,fscal,fix3);
2372             fiy3             = _mm_macc_pd(dy32,fscal,fiy3);
2373             fiz3             = _mm_macc_pd(dz32,fscal,fiz3);
2374
2375             fjx2             = _mm_macc_pd(dx32,fscal,fjx2);
2376             fjy2             = _mm_macc_pd(dy32,fscal,fjy2);
2377             fjz2             = _mm_macc_pd(dz32,fscal,fjz2);
2378
2379             /**************************
2380              * CALCULATE INTERACTIONS *
2381              **************************/
2382
2383             r33              = _mm_mul_pd(rsq33,rinv33);
2384
2385             /* EWALD ELECTROSTATICS */
2386
2387             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2388             ewrt             = _mm_mul_pd(r33,ewtabscale);
2389             ewitab           = _mm_cvttpd_epi32(ewrt);
2390 #ifdef __XOP__
2391             eweps            = _mm_frcz_pd(ewrt);
2392 #else
2393             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2394 #endif
2395             twoeweps         = _mm_add_pd(eweps,eweps);
2396             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2397             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2398             felec            = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
2399
2400             fscal            = felec;
2401
2402             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2403
2404             /* Update vectorial force */
2405             fix3             = _mm_macc_pd(dx33,fscal,fix3);
2406             fiy3             = _mm_macc_pd(dy33,fscal,fiy3);
2407             fiz3             = _mm_macc_pd(dz33,fscal,fiz3);
2408
2409             fjx3             = _mm_macc_pd(dx33,fscal,fjx3);
2410             fjy3             = _mm_macc_pd(dy33,fscal,fjy3);
2411             fjz3             = _mm_macc_pd(dz33,fscal,fjz3);
2412
2413             gmx_mm_decrement_4rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
2414
2415             /* Inner loop uses 405 flops */
2416         }
2417
2418         /* End of innermost loop */
2419
2420         gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
2421                                               f+i_coord_offset,fshift+i_shift_offset);
2422
2423         /* Increment number of inner iterations */
2424         inneriter                  += j_index_end - j_index_start;
2425
2426         /* Outer loop uses 24 flops */
2427     }
2428
2429     /* Increment number of outer iterations */
2430     outeriter        += nri;
2431
2432     /* Update outer/inner flops */
2433
2434     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*24 + inneriter*405);
2435 }