src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_double/nb_kernel_ElecEw_VdwNone_GeomP1P1_avx_256_double.c

   1 /*
   2  * Note: this file was generated by the Gromacs avx_256_double kernel generator.
   3  *
   4  *                This source code is part of
   5  *
   6  *                 G   R   O   M   A   C   S
   7  *
   8  * Copyright (c) 2001-2012, The GROMACS Development Team
   9  *
  10  * Gromacs is a library for molecular simulation and trajectory analysis,
  11  * written by Erik Lindahl, David van der Spoel, Berk Hess, and others - for
  12  * a full list of developers and information, check out http://www.gromacs.org
  13  *
  14  * This program is free software; you can redistribute it and/or modify it under
  15  * the terms of the GNU Lesser General Public License as published by the Free
  16  * Software Foundation; either version 2 of the License, or (at your option) any
  17  * later version.
  18  *
  19  * To help fund GROMACS development, we humbly ask that you cite
  20  * the papers people have written on it - you can find them on the website.
  21  */
  22 #ifdef HAVE_CONFIG_H
  23 #include <config.h>
  24 #endif
  25
  26 #include <math.h>
  27
  28 #include "../nb_kernel.h"
  29 #include "types/simple.h"
  30 #include "vec.h"
  31 #include "nrnb.h"
  32
  33 #include "gmx_math_x86_avx_256_double.h"
  34 #include "kernelutil_x86_avx_256_double.h"
  35
  36 /*
  37  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwNone_GeomP1P1_VF_avx_256_double
  38  * Electrostatics interaction: Ewald
  39  * VdW interaction:            None
  40  * Geometry:                   Particle-Particle
  41  * Calculate force/pot:        PotentialAndForce
  42  */
  43 void
  44 nb_kernel_ElecEw_VdwNone_GeomP1P1_VF_avx_256_double
  45                     (t_nblist * gmx_restrict                nlist,
  46                      rvec * gmx_restrict                    xx,
  47                      rvec * gmx_restrict                    ff,
  48                      t_forcerec * gmx_restrict              fr,
  49                      t_mdatoms * gmx_restrict               mdatoms,
  50                      nb_kernel_data_t * gmx_restrict        kernel_data,
  51                      t_nrnb * gmx_restrict                  nrnb)
  52 {
  53     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  54      * just 0 for non-waters.
  55      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
  56      * jnr indices corresponding to data put in the four positions in the SIMD register.
  57      */
  58     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  59     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  60     int              jnrA,jnrB,jnrC,jnrD;
  61     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  62     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  63     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  64     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  65     real             rcutoff_scalar;
  66     real             *shiftvec,*fshift,*x,*f;
  67     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  68     real             scratch[4*DIM];
  69     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  70     real *           vdwioffsetptr0;
  71     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  72     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  73     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  74     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  75     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
  76     real             *charge;
  77     __m128i          ewitab;
  78     __m256d          ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
  79     __m256d          beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
  80     real             *ewtab;
  81     __m256d          dummy_mask,cutoff_mask;
  82     __m128           tmpmask0,tmpmask1;
  83     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
  84     __m256d          one     = _mm256_set1_pd(1.0);
  85     __m256d          two     = _mm256_set1_pd(2.0);
  86     x                = xx[0];
  87     f                = ff[0];
  88
  89     nri              = nlist->nri;
  90     iinr             = nlist->iinr;
  91     jindex           = nlist->jindex;
  92     jjnr             = nlist->jjnr;
  93     shiftidx         = nlist->shift;
  94     gid              = nlist->gid;
  95     shiftvec         = fr->shift_vec[0];
  96     fshift           = fr->fshift[0];
  97     facel            = _mm256_set1_pd(fr->epsfac);
  98     charge           = mdatoms->chargeA;
  99
 100     sh_ewald         = _mm256_set1_pd(fr->ic->sh_ewald);
 101     beta             = _mm256_set1_pd(fr->ic->ewaldcoeff);
 102     beta2            = _mm256_mul_pd(beta,beta);
 103     beta3            = _mm256_mul_pd(beta,beta2);
 104
 105     ewtab            = fr->ic->tabq_coul_FDV0;
 106     ewtabscale       = _mm256_set1_pd(fr->ic->tabq_scale);
 107     ewtabhalfspace   = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
 108
 109     /* Avoid stupid compiler warnings */
 110     jnrA = jnrB = jnrC = jnrD = 0;
 111     j_coord_offsetA = 0;
 112     j_coord_offsetB = 0;
 113     j_coord_offsetC = 0;
 114     j_coord_offsetD = 0;
 115
 116     outeriter        = 0;
 117     inneriter        = 0;
 118
 119     for(iidx=0;iidx<4*DIM;iidx++)
 120     {
 121         scratch[iidx] = 0.0;
 122     }
 123
 124     /* Start outer loop over neighborlists */
 125     for(iidx=0; iidx<nri; iidx++)
 126     {
 127         /* Load shift vector for this list */
 128         i_shift_offset   = DIM*shiftidx[iidx];
 129
 130         /* Load limits for loop over neighbors */
 131         j_index_start    = jindex[iidx];
 132         j_index_end      = jindex[iidx+1];
 133
 134         /* Get outer coordinate index */
 135         inr              = iinr[iidx];
 136         i_coord_offset   = DIM*inr;
 137
 138         /* Load i particle coords and add shift vector */
 139         gmx_mm256_load_shift_and_1rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 140
 141         fix0             = _mm256_setzero_pd();
 142         fiy0             = _mm256_setzero_pd();
 143         fiz0             = _mm256_setzero_pd();
 144
 145         /* Load parameters for i particles */
 146         iq0              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+0]));
 147
 148         /* Reset potential sums */
 149         velecsum         = _mm256_setzero_pd();
 150
 151         /* Start inner kernel loop */
 152         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 153         {
 154
 155             /* Get j neighbor index, and coordinate index */
 156             jnrA             = jjnr[jidx];
 157             jnrB             = jjnr[jidx+1];
 158             jnrC             = jjnr[jidx+2];
 159             jnrD             = jjnr[jidx+3];
 160             j_coord_offsetA  = DIM*jnrA;
 161             j_coord_offsetB  = DIM*jnrB;
 162             j_coord_offsetC  = DIM*jnrC;
 163             j_coord_offsetD  = DIM*jnrD;
 164
 165             /* load j atom coordinates */
 166             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 167                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 168                                                  &jx0,&jy0,&jz0);
 169
 170             /* Calculate displacement vector */
 171             dx00             = _mm256_sub_pd(ix0,jx0);
 172             dy00             = _mm256_sub_pd(iy0,jy0);
 173             dz00             = _mm256_sub_pd(iz0,jz0);
 174
 175             /* Calculate squared distance and things based on it */
 176             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 177
 178             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
 179
 180             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 181
 182             /* Load parameters for j particles */
 183             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 184                                                                  charge+jnrC+0,charge+jnrD+0);
 185
 186             /**************************
 187              * CALCULATE INTERACTIONS *
 188              **************************/
 189
 190             r00              = _mm256_mul_pd(rsq00,rinv00);
 191
 192             /* Compute parameters for interactions between i and j atoms */
 193             qq00             = _mm256_mul_pd(iq0,jq0);
 194
 195             /* EWALD ELECTROSTATICS */
 196
 197             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 198             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 199             ewitab           = _mm256_cvttpd_epi32(ewrt);
 200             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 201             ewitab           = _mm_slli_epi32(ewitab,2);
 202             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 203             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 204             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 205             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 206             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 207             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 208             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 209             velec            = _mm256_mul_pd(qq00,_mm256_sub_pd(rinv00,velec));
 210             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 211
 212             /* Update potential sum for this i atom from the interaction with this j atom. */
 213             velecsum         = _mm256_add_pd(velecsum,velec);
 214
 215             fscal            = felec;
 216
 217             /* Calculate temporary vectorial force */
 218             tx               = _mm256_mul_pd(fscal,dx00);
 219             ty               = _mm256_mul_pd(fscal,dy00);
 220             tz               = _mm256_mul_pd(fscal,dz00);
 221
 222             /* Update vectorial force */
 223             fix0             = _mm256_add_pd(fix0,tx);
 224             fiy0             = _mm256_add_pd(fiy0,ty);
 225             fiz0             = _mm256_add_pd(fiz0,tz);
 226
 227             fjptrA             = f+j_coord_offsetA;
 228             fjptrB             = f+j_coord_offsetB;
 229             fjptrC             = f+j_coord_offsetC;
 230             fjptrD             = f+j_coord_offsetD;
 231             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
 232
 233             /* Inner loop uses 41 flops */
 234         }
 235
 236         if(jidx<j_index_end)
 237         {
 238
 239             /* Get j neighbor index, and coordinate index */
 240             jnrlistA         = jjnr[jidx];
 241             jnrlistB         = jjnr[jidx+1];
 242             jnrlistC         = jjnr[jidx+2];
 243             jnrlistD         = jjnr[jidx+3];
 244             /* Sign of each element will be negative for non-real atoms.
 245              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 246              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 247              */
 248             tmpmask0 = gmx_mm_castsi128_pd(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 249
 250             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 251             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 252             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 253
 254             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 255             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 256             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 257             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 258             j_coord_offsetA  = DIM*jnrA;
 259             j_coord_offsetB  = DIM*jnrB;
 260             j_coord_offsetC  = DIM*jnrC;
 261             j_coord_offsetD  = DIM*jnrD;
 262
 263             /* load j atom coordinates */
 264             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 265                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 266                                                  &jx0,&jy0,&jz0);
 267
 268             /* Calculate displacement vector */
 269             dx00             = _mm256_sub_pd(ix0,jx0);
 270             dy00             = _mm256_sub_pd(iy0,jy0);
 271             dz00             = _mm256_sub_pd(iz0,jz0);
 272
 273             /* Calculate squared distance and things based on it */
 274             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 275
 276             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
 277
 278             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 279
 280             /* Load parameters for j particles */
 281             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 282                                                                  charge+jnrC+0,charge+jnrD+0);
 283
 284             /**************************
 285              * CALCULATE INTERACTIONS *
 286              **************************/
 287
 288             r00              = _mm256_mul_pd(rsq00,rinv00);
 289             r00              = _mm256_andnot_pd(dummy_mask,r00);
 290
 291             /* Compute parameters for interactions between i and j atoms */
 292             qq00             = _mm256_mul_pd(iq0,jq0);
 293
 294             /* EWALD ELECTROSTATICS */
 295
 296             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 297             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 298             ewitab           = _mm256_cvttpd_epi32(ewrt);
 299             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 300             ewitab           = _mm_slli_epi32(ewitab,2);
 301             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 302             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 303             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 304             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 305             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 306             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 307             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 308             velec            = _mm256_mul_pd(qq00,_mm256_sub_pd(rinv00,velec));
 309             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 310
 311             /* Update potential sum for this i atom from the interaction with this j atom. */
 312             velec            = _mm256_andnot_pd(dummy_mask,velec);
 313             velecsum         = _mm256_add_pd(velecsum,velec);
 314
 315             fscal            = felec;
 316
 317             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 318
 319             /* Calculate temporary vectorial force */
 320             tx               = _mm256_mul_pd(fscal,dx00);
 321             ty               = _mm256_mul_pd(fscal,dy00);
 322             tz               = _mm256_mul_pd(fscal,dz00);
 323
 324             /* Update vectorial force */
 325             fix0             = _mm256_add_pd(fix0,tx);
 326             fiy0             = _mm256_add_pd(fiy0,ty);
 327             fiz0             = _mm256_add_pd(fiz0,tz);
 328
 329             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 330             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 331             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 332             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 333             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
 334
 335             /* Inner loop uses 42 flops */
 336         }
 337
 338         /* End of innermost loop */
 339
 340         gmx_mm256_update_iforce_1atom_swizzle_pd(fix0,fiy0,fiz0,
 341                                                  f+i_coord_offset,fshift+i_shift_offset);
 342
 343         ggid                        = gid[iidx];
 344         /* Update potential energies */
 345         gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
 346
 347         /* Increment number of inner iterations */
 348         inneriter                  += j_index_end - j_index_start;
 349
 350         /* Outer loop uses 8 flops */
 351     }
 352
 353     /* Increment number of outer iterations */
 354     outeriter        += nri;
 355
 356     /* Update outer/inner flops */
 357
 358     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VF,outeriter*8 + inneriter*42);
 359 }
 360 /*
 361  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwNone_GeomP1P1_F_avx_256_double
 362  * Electrostatics interaction: Ewald
 363  * VdW interaction:            None
 364  * Geometry:                   Particle-Particle
 365  * Calculate force/pot:        Force
 366  */
 367 void
 368 nb_kernel_ElecEw_VdwNone_GeomP1P1_F_avx_256_double
 369                     (t_nblist * gmx_restrict                nlist,
 370                      rvec * gmx_restrict                    xx,
 371                      rvec * gmx_restrict                    ff,
 372                      t_forcerec * gmx_restrict              fr,
 373                      t_mdatoms * gmx_restrict               mdatoms,
 374                      nb_kernel_data_t * gmx_restrict        kernel_data,
 375                      t_nrnb * gmx_restrict                  nrnb)
 376 {
 377     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 378      * just 0 for non-waters.
 379      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
 380      * jnr indices corresponding to data put in the four positions in the SIMD register.
 381      */
 382     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 383     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 384     int              jnrA,jnrB,jnrC,jnrD;
 385     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 386     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 387     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 388     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 389     real             rcutoff_scalar;
 390     real             *shiftvec,*fshift,*x,*f;
 391     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 392     real             scratch[4*DIM];
 393     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 394     real *           vdwioffsetptr0;
 395     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 396     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 397     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 398     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 399     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
 400     real             *charge;
 401     __m128i          ewitab;
 402     __m256d          ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 403     __m256d          beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 404     real             *ewtab;
 405     __m256d          dummy_mask,cutoff_mask;
 406     __m128           tmpmask0,tmpmask1;
 407     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 408     __m256d          one     = _mm256_set1_pd(1.0);
 409     __m256d          two     = _mm256_set1_pd(2.0);
 410     x                = xx[0];
 411     f                = ff[0];
 412
 413     nri              = nlist->nri;
 414     iinr             = nlist->iinr;
 415     jindex           = nlist->jindex;
 416     jjnr             = nlist->jjnr;
 417     shiftidx         = nlist->shift;
 418     gid              = nlist->gid;
 419     shiftvec         = fr->shift_vec[0];
 420     fshift           = fr->fshift[0];
 421     facel            = _mm256_set1_pd(fr->epsfac);
 422     charge           = mdatoms->chargeA;
 423
 424     sh_ewald         = _mm256_set1_pd(fr->ic->sh_ewald);
 425     beta             = _mm256_set1_pd(fr->ic->ewaldcoeff);
 426     beta2            = _mm256_mul_pd(beta,beta);
 427     beta3            = _mm256_mul_pd(beta,beta2);
 428
 429     ewtab            = fr->ic->tabq_coul_F;
 430     ewtabscale       = _mm256_set1_pd(fr->ic->tabq_scale);
 431     ewtabhalfspace   = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
 432
 433     /* Avoid stupid compiler warnings */
 434     jnrA = jnrB = jnrC = jnrD = 0;
 435     j_coord_offsetA = 0;
 436     j_coord_offsetB = 0;
 437     j_coord_offsetC = 0;
 438     j_coord_offsetD = 0;
 439
 440     outeriter        = 0;
 441     inneriter        = 0;
 442
 443     for(iidx=0;iidx<4*DIM;iidx++)
 444     {
 445         scratch[iidx] = 0.0;
 446     }
 447
 448     /* Start outer loop over neighborlists */
 449     for(iidx=0; iidx<nri; iidx++)
 450     {
 451         /* Load shift vector for this list */
 452         i_shift_offset   = DIM*shiftidx[iidx];
 453
 454         /* Load limits for loop over neighbors */
 455         j_index_start    = jindex[iidx];
 456         j_index_end      = jindex[iidx+1];
 457
 458         /* Get outer coordinate index */
 459         inr              = iinr[iidx];
 460         i_coord_offset   = DIM*inr;
 461
 462         /* Load i particle coords and add shift vector */
 463         gmx_mm256_load_shift_and_1rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 464
 465         fix0             = _mm256_setzero_pd();
 466         fiy0             = _mm256_setzero_pd();
 467         fiz0             = _mm256_setzero_pd();
 468
 469         /* Load parameters for i particles */
 470         iq0              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+0]));
 471
 472         /* Start inner kernel loop */
 473         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 474         {
 475
 476             /* Get j neighbor index, and coordinate index */
 477             jnrA             = jjnr[jidx];
 478             jnrB             = jjnr[jidx+1];
 479             jnrC             = jjnr[jidx+2];
 480             jnrD             = jjnr[jidx+3];
 481             j_coord_offsetA  = DIM*jnrA;
 482             j_coord_offsetB  = DIM*jnrB;
 483             j_coord_offsetC  = DIM*jnrC;
 484             j_coord_offsetD  = DIM*jnrD;
 485
 486             /* load j atom coordinates */
 487             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 488                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 489                                                  &jx0,&jy0,&jz0);
 490
 491             /* Calculate displacement vector */
 492             dx00             = _mm256_sub_pd(ix0,jx0);
 493             dy00             = _mm256_sub_pd(iy0,jy0);
 494             dz00             = _mm256_sub_pd(iz0,jz0);
 495
 496             /* Calculate squared distance and things based on it */
 497             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 498
 499             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
 500
 501             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 502
 503             /* Load parameters for j particles */
 504             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 505                                                                  charge+jnrC+0,charge+jnrD+0);
 506
 507             /**************************
 508              * CALCULATE INTERACTIONS *
 509              **************************/
 510
 511             r00              = _mm256_mul_pd(rsq00,rinv00);
 512
 513             /* Compute parameters for interactions between i and j atoms */
 514             qq00             = _mm256_mul_pd(iq0,jq0);
 515
 516             /* EWALD ELECTROSTATICS */
 517
 518             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 519             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 520             ewitab           = _mm256_cvttpd_epi32(ewrt);
 521             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 522             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
 523                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
 524                                             &ewtabF,&ewtabFn);
 525             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
 526             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 527
 528             fscal            = felec;
 529
 530             /* Calculate temporary vectorial force */
 531             tx               = _mm256_mul_pd(fscal,dx00);
 532             ty               = _mm256_mul_pd(fscal,dy00);
 533             tz               = _mm256_mul_pd(fscal,dz00);
 534
 535             /* Update vectorial force */
 536             fix0             = _mm256_add_pd(fix0,tx);
 537             fiy0             = _mm256_add_pd(fiy0,ty);
 538             fiz0             = _mm256_add_pd(fiz0,tz);
 539
 540             fjptrA             = f+j_coord_offsetA;
 541             fjptrB             = f+j_coord_offsetB;
 542             fjptrC             = f+j_coord_offsetC;
 543             fjptrD             = f+j_coord_offsetD;
 544             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
 545
 546             /* Inner loop uses 36 flops */
 547         }
 548
 549         if(jidx<j_index_end)
 550         {
 551
 552             /* Get j neighbor index, and coordinate index */
 553             jnrlistA         = jjnr[jidx];
 554             jnrlistB         = jjnr[jidx+1];
 555             jnrlistC         = jjnr[jidx+2];
 556             jnrlistD         = jjnr[jidx+3];
 557             /* Sign of each element will be negative for non-real atoms.
 558              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 559              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 560              */
 561             tmpmask0 = gmx_mm_castsi128_pd(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 562
 563             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 564             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 565             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 566
 567             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 568             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 569             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 570             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 571             j_coord_offsetA  = DIM*jnrA;
 572             j_coord_offsetB  = DIM*jnrB;
 573             j_coord_offsetC  = DIM*jnrC;
 574             j_coord_offsetD  = DIM*jnrD;
 575
 576             /* load j atom coordinates */
 577             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 578                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 579                                                  &jx0,&jy0,&jz0);
 580
 581             /* Calculate displacement vector */
 582             dx00             = _mm256_sub_pd(ix0,jx0);
 583             dy00             = _mm256_sub_pd(iy0,jy0);
 584             dz00             = _mm256_sub_pd(iz0,jz0);
 585
 586             /* Calculate squared distance and things based on it */
 587             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 588
 589             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
 590
 591             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 592
 593             /* Load parameters for j particles */
 594             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 595                                                                  charge+jnrC+0,charge+jnrD+0);
 596
 597             /**************************
 598              * CALCULATE INTERACTIONS *
 599              **************************/
 600
 601             r00              = _mm256_mul_pd(rsq00,rinv00);
 602             r00              = _mm256_andnot_pd(dummy_mask,r00);
 603
 604             /* Compute parameters for interactions between i and j atoms */
 605             qq00             = _mm256_mul_pd(iq0,jq0);
 606
 607             /* EWALD ELECTROSTATICS */
 608
 609             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 610             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 611             ewitab           = _mm256_cvttpd_epi32(ewrt);
 612             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 613             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
 614                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
 615                                             &ewtabF,&ewtabFn);
 616             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
 617             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 618
 619             fscal            = felec;
 620
 621             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 622
 623             /* Calculate temporary vectorial force */
 624             tx               = _mm256_mul_pd(fscal,dx00);
 625             ty               = _mm256_mul_pd(fscal,dy00);
 626             tz               = _mm256_mul_pd(fscal,dz00);
 627
 628             /* Update vectorial force */
 629             fix0             = _mm256_add_pd(fix0,tx);
 630             fiy0             = _mm256_add_pd(fiy0,ty);
 631             fiz0             = _mm256_add_pd(fiz0,tz);
 632
 633             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 634             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 635             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 636             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 637             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
 638
 639             /* Inner loop uses 37 flops */
 640         }
 641
 642         /* End of innermost loop */
 643
 644         gmx_mm256_update_iforce_1atom_swizzle_pd(fix0,fiy0,fiz0,
 645                                                  f+i_coord_offset,fshift+i_shift_offset);
 646
 647         /* Increment number of inner iterations */
 648         inneriter                  += j_index_end - j_index_start;
 649
 650         /* Outer loop uses 7 flops */
 651     }
 652
 653     /* Increment number of outer iterations */
 654     outeriter        += nri;
 655
 656     /* Update outer/inner flops */
 657
 658     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_F,outeriter*7 + inneriter*37);
 659 }