src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_double/nb_kernel_ElecCoul_VdwNone_GeomW4P1_avx_256_double.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014,2015,2017, 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_256_double kernel generator.
  37  */
  38 #include "gmxpre.h"
  39
  40 #include "config.h"
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "gromacs/gmxlib/nrnb.h"
  46
  47 #include "kernelutil_x86_avx_256_double.h"
  48
  49 /*
  50  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwNone_GeomW4P1_VF_avx_256_double
  51  * Electrostatics interaction: Coulomb
  52  * VdW interaction:            None
  53  * Geometry:                   Water4-Particle
  54  * Calculate force/pot:        PotentialAndForce
  55  */
  56 void
  57 nb_kernel_ElecCoul_VdwNone_GeomW4P1_VF_avx_256_double
  58                     (t_nblist                    * gmx_restrict       nlist,
  59                      rvec                        * gmx_restrict          xx,
  60                      rvec                        * gmx_restrict          ff,
  61                      struct t_forcerec           * gmx_restrict          fr,
  62                      t_mdatoms                   * gmx_restrict     mdatoms,
  63                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  64                      t_nrnb                      * gmx_restrict        nrnb)
  65 {
  66     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  67      * just 0 for non-waters.
  68      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
  69      * jnr indices corresponding to data put in the four positions in the SIMD register.
  70      */
  71     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  72     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  73     int              jnrA,jnrB,jnrC,jnrD;
  74     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  75     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  76     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  77     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  78     real             rcutoff_scalar;
  79     real             *shiftvec,*fshift,*x,*f;
  80     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  81     real             scratch[4*DIM];
  82     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  83     real *           vdwioffsetptr1;
  84     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  85     real *           vdwioffsetptr2;
  86     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  87     real *           vdwioffsetptr3;
  88     __m256d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  89     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  90     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  91     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  92     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  93     __m256d          dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
  94     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
  95     real             *charge;
  96     __m256d          dummy_mask,cutoff_mask;
  97     __m128           tmpmask0,tmpmask1;
  98     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
  99     __m256d          one     = _mm256_set1_pd(1.0);
 100     __m256d          two     = _mm256_set1_pd(2.0);
 101     x                = xx[0];
 102     f                = ff[0];
 103
 104     nri              = nlist->nri;
 105     iinr             = nlist->iinr;
 106     jindex           = nlist->jindex;
 107     jjnr             = nlist->jjnr;
 108     shiftidx         = nlist->shift;
 109     gid              = nlist->gid;
 110     shiftvec         = fr->shift_vec[0];
 111     fshift           = fr->fshift[0];
 112     facel            = _mm256_set1_pd(fr->ic->epsfac);
 113     charge           = mdatoms->chargeA;
 114
 115     /* Setup water-specific parameters */
 116     inr              = nlist->iinr[0];
 117     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 118     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 119     iq3              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
 120
 121     /* Avoid stupid compiler warnings */
 122     jnrA = jnrB = jnrC = jnrD = 0;
 123     j_coord_offsetA = 0;
 124     j_coord_offsetB = 0;
 125     j_coord_offsetC = 0;
 126     j_coord_offsetD = 0;
 127
 128     outeriter        = 0;
 129     inneriter        = 0;
 130
 131     for(iidx=0;iidx<4*DIM;iidx++)
 132     {
 133         scratch[iidx] = 0.0;
 134     }
 135
 136     /* Start outer loop over neighborlists */
 137     for(iidx=0; iidx<nri; iidx++)
 138     {
 139         /* Load shift vector for this list */
 140         i_shift_offset   = DIM*shiftidx[iidx];
 141
 142         /* Load limits for loop over neighbors */
 143         j_index_start    = jindex[iidx];
 144         j_index_end      = jindex[iidx+1];
 145
 146         /* Get outer coordinate index */
 147         inr              = iinr[iidx];
 148         i_coord_offset   = DIM*inr;
 149
 150         /* Load i particle coords and add shift vector */
 151         gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset+DIM,
 152                                                     &ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 153
 154         fix1             = _mm256_setzero_pd();
 155         fiy1             = _mm256_setzero_pd();
 156         fiz1             = _mm256_setzero_pd();
 157         fix2             = _mm256_setzero_pd();
 158         fiy2             = _mm256_setzero_pd();
 159         fiz2             = _mm256_setzero_pd();
 160         fix3             = _mm256_setzero_pd();
 161         fiy3             = _mm256_setzero_pd();
 162         fiz3             = _mm256_setzero_pd();
 163
 164         /* Reset potential sums */
 165         velecsum         = _mm256_setzero_pd();
 166
 167         /* Start inner kernel loop */
 168         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 169         {
 170
 171             /* Get j neighbor index, and coordinate index */
 172             jnrA             = jjnr[jidx];
 173             jnrB             = jjnr[jidx+1];
 174             jnrC             = jjnr[jidx+2];
 175             jnrD             = jjnr[jidx+3];
 176             j_coord_offsetA  = DIM*jnrA;
 177             j_coord_offsetB  = DIM*jnrB;
 178             j_coord_offsetC  = DIM*jnrC;
 179             j_coord_offsetD  = DIM*jnrD;
 180
 181             /* load j atom coordinates */
 182             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 183                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 184                                                  &jx0,&jy0,&jz0);
 185
 186             /* Calculate displacement vector */
 187             dx10             = _mm256_sub_pd(ix1,jx0);
 188             dy10             = _mm256_sub_pd(iy1,jy0);
 189             dz10             = _mm256_sub_pd(iz1,jz0);
 190             dx20             = _mm256_sub_pd(ix2,jx0);
 191             dy20             = _mm256_sub_pd(iy2,jy0);
 192             dz20             = _mm256_sub_pd(iz2,jz0);
 193             dx30             = _mm256_sub_pd(ix3,jx0);
 194             dy30             = _mm256_sub_pd(iy3,jy0);
 195             dz30             = _mm256_sub_pd(iz3,jz0);
 196
 197             /* Calculate squared distance and things based on it */
 198             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 199             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 200             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 201
 202             rinv10           = avx256_invsqrt_d(rsq10);
 203             rinv20           = avx256_invsqrt_d(rsq20);
 204             rinv30           = avx256_invsqrt_d(rsq30);
 205
 206             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 207             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 208             rinvsq30         = _mm256_mul_pd(rinv30,rinv30);
 209
 210             /* Load parameters for j particles */
 211             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 212                                                                  charge+jnrC+0,charge+jnrD+0);
 213
 214             fjx0             = _mm256_setzero_pd();
 215             fjy0             = _mm256_setzero_pd();
 216             fjz0             = _mm256_setzero_pd();
 217
 218             /**************************
 219              * CALCULATE INTERACTIONS *
 220              **************************/
 221
 222             /* Compute parameters for interactions between i and j atoms */
 223             qq10             = _mm256_mul_pd(iq1,jq0);
 224
 225             /* COULOMB ELECTROSTATICS */
 226             velec            = _mm256_mul_pd(qq10,rinv10);
 227             felec            = _mm256_mul_pd(velec,rinvsq10);
 228
 229             /* Update potential sum for this i atom from the interaction with this j atom. */
 230             velecsum         = _mm256_add_pd(velecsum,velec);
 231
 232             fscal            = felec;
 233
 234             /* Calculate temporary vectorial force */
 235             tx               = _mm256_mul_pd(fscal,dx10);
 236             ty               = _mm256_mul_pd(fscal,dy10);
 237             tz               = _mm256_mul_pd(fscal,dz10);
 238
 239             /* Update vectorial force */
 240             fix1             = _mm256_add_pd(fix1,tx);
 241             fiy1             = _mm256_add_pd(fiy1,ty);
 242             fiz1             = _mm256_add_pd(fiz1,tz);
 243
 244             fjx0             = _mm256_add_pd(fjx0,tx);
 245             fjy0             = _mm256_add_pd(fjy0,ty);
 246             fjz0             = _mm256_add_pd(fjz0,tz);
 247
 248             /**************************
 249              * CALCULATE INTERACTIONS *
 250              **************************/
 251
 252             /* Compute parameters for interactions between i and j atoms */
 253             qq20             = _mm256_mul_pd(iq2,jq0);
 254
 255             /* COULOMB ELECTROSTATICS */
 256             velec            = _mm256_mul_pd(qq20,rinv20);
 257             felec            = _mm256_mul_pd(velec,rinvsq20);
 258
 259             /* Update potential sum for this i atom from the interaction with this j atom. */
 260             velecsum         = _mm256_add_pd(velecsum,velec);
 261
 262             fscal            = felec;
 263
 264             /* Calculate temporary vectorial force */
 265             tx               = _mm256_mul_pd(fscal,dx20);
 266             ty               = _mm256_mul_pd(fscal,dy20);
 267             tz               = _mm256_mul_pd(fscal,dz20);
 268
 269             /* Update vectorial force */
 270             fix2             = _mm256_add_pd(fix2,tx);
 271             fiy2             = _mm256_add_pd(fiy2,ty);
 272             fiz2             = _mm256_add_pd(fiz2,tz);
 273
 274             fjx0             = _mm256_add_pd(fjx0,tx);
 275             fjy0             = _mm256_add_pd(fjy0,ty);
 276             fjz0             = _mm256_add_pd(fjz0,tz);
 277
 278             /**************************
 279              * CALCULATE INTERACTIONS *
 280              **************************/
 281
 282             /* Compute parameters for interactions between i and j atoms */
 283             qq30             = _mm256_mul_pd(iq3,jq0);
 284
 285             /* COULOMB ELECTROSTATICS */
 286             velec            = _mm256_mul_pd(qq30,rinv30);
 287             felec            = _mm256_mul_pd(velec,rinvsq30);
 288
 289             /* Update potential sum for this i atom from the interaction with this j atom. */
 290             velecsum         = _mm256_add_pd(velecsum,velec);
 291
 292             fscal            = felec;
 293
 294             /* Calculate temporary vectorial force */
 295             tx               = _mm256_mul_pd(fscal,dx30);
 296             ty               = _mm256_mul_pd(fscal,dy30);
 297             tz               = _mm256_mul_pd(fscal,dz30);
 298
 299             /* Update vectorial force */
 300             fix3             = _mm256_add_pd(fix3,tx);
 301             fiy3             = _mm256_add_pd(fiy3,ty);
 302             fiz3             = _mm256_add_pd(fiz3,tz);
 303
 304             fjx0             = _mm256_add_pd(fjx0,tx);
 305             fjy0             = _mm256_add_pd(fjy0,ty);
 306             fjz0             = _mm256_add_pd(fjz0,tz);
 307
 308             fjptrA             = f+j_coord_offsetA;
 309             fjptrB             = f+j_coord_offsetB;
 310             fjptrC             = f+j_coord_offsetC;
 311             fjptrD             = f+j_coord_offsetD;
 312
 313             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 314
 315             /* Inner loop uses 84 flops */
 316         }
 317
 318         if(jidx<j_index_end)
 319         {
 320
 321             /* Get j neighbor index, and coordinate index */
 322             jnrlistA         = jjnr[jidx];
 323             jnrlistB         = jjnr[jidx+1];
 324             jnrlistC         = jjnr[jidx+2];
 325             jnrlistD         = jjnr[jidx+3];
 326             /* Sign of each element will be negative for non-real atoms.
 327              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 328              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 329              */
 330             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 331
 332             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 333             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 334             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 335
 336             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 337             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 338             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 339             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 340             j_coord_offsetA  = DIM*jnrA;
 341             j_coord_offsetB  = DIM*jnrB;
 342             j_coord_offsetC  = DIM*jnrC;
 343             j_coord_offsetD  = DIM*jnrD;
 344
 345             /* load j atom coordinates */
 346             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 347                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 348                                                  &jx0,&jy0,&jz0);
 349
 350             /* Calculate displacement vector */
 351             dx10             = _mm256_sub_pd(ix1,jx0);
 352             dy10             = _mm256_sub_pd(iy1,jy0);
 353             dz10             = _mm256_sub_pd(iz1,jz0);
 354             dx20             = _mm256_sub_pd(ix2,jx0);
 355             dy20             = _mm256_sub_pd(iy2,jy0);
 356             dz20             = _mm256_sub_pd(iz2,jz0);
 357             dx30             = _mm256_sub_pd(ix3,jx0);
 358             dy30             = _mm256_sub_pd(iy3,jy0);
 359             dz30             = _mm256_sub_pd(iz3,jz0);
 360
 361             /* Calculate squared distance and things based on it */
 362             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 363             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 364             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 365
 366             rinv10           = avx256_invsqrt_d(rsq10);
 367             rinv20           = avx256_invsqrt_d(rsq20);
 368             rinv30           = avx256_invsqrt_d(rsq30);
 369
 370             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 371             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 372             rinvsq30         = _mm256_mul_pd(rinv30,rinv30);
 373
 374             /* Load parameters for j particles */
 375             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 376                                                                  charge+jnrC+0,charge+jnrD+0);
 377
 378             fjx0             = _mm256_setzero_pd();
 379             fjy0             = _mm256_setzero_pd();
 380             fjz0             = _mm256_setzero_pd();
 381
 382             /**************************
 383              * CALCULATE INTERACTIONS *
 384              **************************/
 385
 386             /* Compute parameters for interactions between i and j atoms */
 387             qq10             = _mm256_mul_pd(iq1,jq0);
 388
 389             /* COULOMB ELECTROSTATICS */
 390             velec            = _mm256_mul_pd(qq10,rinv10);
 391             felec            = _mm256_mul_pd(velec,rinvsq10);
 392
 393             /* Update potential sum for this i atom from the interaction with this j atom. */
 394             velec            = _mm256_andnot_pd(dummy_mask,velec);
 395             velecsum         = _mm256_add_pd(velecsum,velec);
 396
 397             fscal            = felec;
 398
 399             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 400
 401             /* Calculate temporary vectorial force */
 402             tx               = _mm256_mul_pd(fscal,dx10);
 403             ty               = _mm256_mul_pd(fscal,dy10);
 404             tz               = _mm256_mul_pd(fscal,dz10);
 405
 406             /* Update vectorial force */
 407             fix1             = _mm256_add_pd(fix1,tx);
 408             fiy1             = _mm256_add_pd(fiy1,ty);
 409             fiz1             = _mm256_add_pd(fiz1,tz);
 410
 411             fjx0             = _mm256_add_pd(fjx0,tx);
 412             fjy0             = _mm256_add_pd(fjy0,ty);
 413             fjz0             = _mm256_add_pd(fjz0,tz);
 414
 415             /**************************
 416              * CALCULATE INTERACTIONS *
 417              **************************/
 418
 419             /* Compute parameters for interactions between i and j atoms */
 420             qq20             = _mm256_mul_pd(iq2,jq0);
 421
 422             /* COULOMB ELECTROSTATICS */
 423             velec            = _mm256_mul_pd(qq20,rinv20);
 424             felec            = _mm256_mul_pd(velec,rinvsq20);
 425
 426             /* Update potential sum for this i atom from the interaction with this j atom. */
 427             velec            = _mm256_andnot_pd(dummy_mask,velec);
 428             velecsum         = _mm256_add_pd(velecsum,velec);
 429
 430             fscal            = felec;
 431
 432             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 433
 434             /* Calculate temporary vectorial force */
 435             tx               = _mm256_mul_pd(fscal,dx20);
 436             ty               = _mm256_mul_pd(fscal,dy20);
 437             tz               = _mm256_mul_pd(fscal,dz20);
 438
 439             /* Update vectorial force */
 440             fix2             = _mm256_add_pd(fix2,tx);
 441             fiy2             = _mm256_add_pd(fiy2,ty);
 442             fiz2             = _mm256_add_pd(fiz2,tz);
 443
 444             fjx0             = _mm256_add_pd(fjx0,tx);
 445             fjy0             = _mm256_add_pd(fjy0,ty);
 446             fjz0             = _mm256_add_pd(fjz0,tz);
 447
 448             /**************************
 449              * CALCULATE INTERACTIONS *
 450              **************************/
 451
 452             /* Compute parameters for interactions between i and j atoms */
 453             qq30             = _mm256_mul_pd(iq3,jq0);
 454
 455             /* COULOMB ELECTROSTATICS */
 456             velec            = _mm256_mul_pd(qq30,rinv30);
 457             felec            = _mm256_mul_pd(velec,rinvsq30);
 458
 459             /* Update potential sum for this i atom from the interaction with this j atom. */
 460             velec            = _mm256_andnot_pd(dummy_mask,velec);
 461             velecsum         = _mm256_add_pd(velecsum,velec);
 462
 463             fscal            = felec;
 464
 465             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 466
 467             /* Calculate temporary vectorial force */
 468             tx               = _mm256_mul_pd(fscal,dx30);
 469             ty               = _mm256_mul_pd(fscal,dy30);
 470             tz               = _mm256_mul_pd(fscal,dz30);
 471
 472             /* Update vectorial force */
 473             fix3             = _mm256_add_pd(fix3,tx);
 474             fiy3             = _mm256_add_pd(fiy3,ty);
 475             fiz3             = _mm256_add_pd(fiz3,tz);
 476
 477             fjx0             = _mm256_add_pd(fjx0,tx);
 478             fjy0             = _mm256_add_pd(fjy0,ty);
 479             fjz0             = _mm256_add_pd(fjz0,tz);
 480
 481             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 482             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 483             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 484             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 485
 486             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 487
 488             /* Inner loop uses 84 flops */
 489         }
 490
 491         /* End of innermost loop */
 492
 493         gmx_mm256_update_iforce_3atom_swizzle_pd(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 494                                                  f+i_coord_offset+DIM,fshift+i_shift_offset);
 495
 496         ggid                        = gid[iidx];
 497         /* Update potential energies */
 498         gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
 499
 500         /* Increment number of inner iterations */
 501         inneriter                  += j_index_end - j_index_start;
 502
 503         /* Outer loop uses 19 flops */
 504     }
 505
 506     /* Increment number of outer iterations */
 507     outeriter        += nri;
 508
 509     /* Update outer/inner flops */
 510
 511     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4_VF,outeriter*19 + inneriter*84);
 512 }
 513 /*
 514  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwNone_GeomW4P1_F_avx_256_double
 515  * Electrostatics interaction: Coulomb
 516  * VdW interaction:            None
 517  * Geometry:                   Water4-Particle
 518  * Calculate force/pot:        Force
 519  */
 520 void
 521 nb_kernel_ElecCoul_VdwNone_GeomW4P1_F_avx_256_double
 522                     (t_nblist                    * gmx_restrict       nlist,
 523                      rvec                        * gmx_restrict          xx,
 524                      rvec                        * gmx_restrict          ff,
 525                      struct t_forcerec           * gmx_restrict          fr,
 526                      t_mdatoms                   * gmx_restrict     mdatoms,
 527                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 528                      t_nrnb                      * gmx_restrict        nrnb)
 529 {
 530     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 531      * just 0 for non-waters.
 532      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
 533      * jnr indices corresponding to data put in the four positions in the SIMD register.
 534      */
 535     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 536     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 537     int              jnrA,jnrB,jnrC,jnrD;
 538     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 539     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 540     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 541     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 542     real             rcutoff_scalar;
 543     real             *shiftvec,*fshift,*x,*f;
 544     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 545     real             scratch[4*DIM];
 546     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 547     real *           vdwioffsetptr1;
 548     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 549     real *           vdwioffsetptr2;
 550     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 551     real *           vdwioffsetptr3;
 552     __m256d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
 553     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 554     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 555     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 556     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 557     __m256d          dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
 558     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
 559     real             *charge;
 560     __m256d          dummy_mask,cutoff_mask;
 561     __m128           tmpmask0,tmpmask1;
 562     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 563     __m256d          one     = _mm256_set1_pd(1.0);
 564     __m256d          two     = _mm256_set1_pd(2.0);
 565     x                = xx[0];
 566     f                = ff[0];
 567
 568     nri              = nlist->nri;
 569     iinr             = nlist->iinr;
 570     jindex           = nlist->jindex;
 571     jjnr             = nlist->jjnr;
 572     shiftidx         = nlist->shift;
 573     gid              = nlist->gid;
 574     shiftvec         = fr->shift_vec[0];
 575     fshift           = fr->fshift[0];
 576     facel            = _mm256_set1_pd(fr->ic->epsfac);
 577     charge           = mdatoms->chargeA;
 578
 579     /* Setup water-specific parameters */
 580     inr              = nlist->iinr[0];
 581     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 582     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 583     iq3              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
 584
 585     /* Avoid stupid compiler warnings */
 586     jnrA = jnrB = jnrC = jnrD = 0;
 587     j_coord_offsetA = 0;
 588     j_coord_offsetB = 0;
 589     j_coord_offsetC = 0;
 590     j_coord_offsetD = 0;
 591
 592     outeriter        = 0;
 593     inneriter        = 0;
 594
 595     for(iidx=0;iidx<4*DIM;iidx++)
 596     {
 597         scratch[iidx] = 0.0;
 598     }
 599
 600     /* Start outer loop over neighborlists */
 601     for(iidx=0; iidx<nri; iidx++)
 602     {
 603         /* Load shift vector for this list */
 604         i_shift_offset   = DIM*shiftidx[iidx];
 605
 606         /* Load limits for loop over neighbors */
 607         j_index_start    = jindex[iidx];
 608         j_index_end      = jindex[iidx+1];
 609
 610         /* Get outer coordinate index */
 611         inr              = iinr[iidx];
 612         i_coord_offset   = DIM*inr;
 613
 614         /* Load i particle coords and add shift vector */
 615         gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset+DIM,
 616                                                     &ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 617
 618         fix1             = _mm256_setzero_pd();
 619         fiy1             = _mm256_setzero_pd();
 620         fiz1             = _mm256_setzero_pd();
 621         fix2             = _mm256_setzero_pd();
 622         fiy2             = _mm256_setzero_pd();
 623         fiz2             = _mm256_setzero_pd();
 624         fix3             = _mm256_setzero_pd();
 625         fiy3             = _mm256_setzero_pd();
 626         fiz3             = _mm256_setzero_pd();
 627
 628         /* Start inner kernel loop */
 629         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 630         {
 631
 632             /* Get j neighbor index, and coordinate index */
 633             jnrA             = jjnr[jidx];
 634             jnrB             = jjnr[jidx+1];
 635             jnrC             = jjnr[jidx+2];
 636             jnrD             = jjnr[jidx+3];
 637             j_coord_offsetA  = DIM*jnrA;
 638             j_coord_offsetB  = DIM*jnrB;
 639             j_coord_offsetC  = DIM*jnrC;
 640             j_coord_offsetD  = DIM*jnrD;
 641
 642             /* load j atom coordinates */
 643             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 644                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 645                                                  &jx0,&jy0,&jz0);
 646
 647             /* Calculate displacement vector */
 648             dx10             = _mm256_sub_pd(ix1,jx0);
 649             dy10             = _mm256_sub_pd(iy1,jy0);
 650             dz10             = _mm256_sub_pd(iz1,jz0);
 651             dx20             = _mm256_sub_pd(ix2,jx0);
 652             dy20             = _mm256_sub_pd(iy2,jy0);
 653             dz20             = _mm256_sub_pd(iz2,jz0);
 654             dx30             = _mm256_sub_pd(ix3,jx0);
 655             dy30             = _mm256_sub_pd(iy3,jy0);
 656             dz30             = _mm256_sub_pd(iz3,jz0);
 657
 658             /* Calculate squared distance and things based on it */
 659             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 660             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 661             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 662
 663             rinv10           = avx256_invsqrt_d(rsq10);
 664             rinv20           = avx256_invsqrt_d(rsq20);
 665             rinv30           = avx256_invsqrt_d(rsq30);
 666
 667             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 668             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 669             rinvsq30         = _mm256_mul_pd(rinv30,rinv30);
 670
 671             /* Load parameters for j particles */
 672             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 673                                                                  charge+jnrC+0,charge+jnrD+0);
 674
 675             fjx0             = _mm256_setzero_pd();
 676             fjy0             = _mm256_setzero_pd();
 677             fjz0             = _mm256_setzero_pd();
 678
 679             /**************************
 680              * CALCULATE INTERACTIONS *
 681              **************************/
 682
 683             /* Compute parameters for interactions between i and j atoms */
 684             qq10             = _mm256_mul_pd(iq1,jq0);
 685
 686             /* COULOMB ELECTROSTATICS */
 687             velec            = _mm256_mul_pd(qq10,rinv10);
 688             felec            = _mm256_mul_pd(velec,rinvsq10);
 689
 690             fscal            = felec;
 691
 692             /* Calculate temporary vectorial force */
 693             tx               = _mm256_mul_pd(fscal,dx10);
 694             ty               = _mm256_mul_pd(fscal,dy10);
 695             tz               = _mm256_mul_pd(fscal,dz10);
 696
 697             /* Update vectorial force */
 698             fix1             = _mm256_add_pd(fix1,tx);
 699             fiy1             = _mm256_add_pd(fiy1,ty);
 700             fiz1             = _mm256_add_pd(fiz1,tz);
 701
 702             fjx0             = _mm256_add_pd(fjx0,tx);
 703             fjy0             = _mm256_add_pd(fjy0,ty);
 704             fjz0             = _mm256_add_pd(fjz0,tz);
 705
 706             /**************************
 707              * CALCULATE INTERACTIONS *
 708              **************************/
 709
 710             /* Compute parameters for interactions between i and j atoms */
 711             qq20             = _mm256_mul_pd(iq2,jq0);
 712
 713             /* COULOMB ELECTROSTATICS */
 714             velec            = _mm256_mul_pd(qq20,rinv20);
 715             felec            = _mm256_mul_pd(velec,rinvsq20);
 716
 717             fscal            = felec;
 718
 719             /* Calculate temporary vectorial force */
 720             tx               = _mm256_mul_pd(fscal,dx20);
 721             ty               = _mm256_mul_pd(fscal,dy20);
 722             tz               = _mm256_mul_pd(fscal,dz20);
 723
 724             /* Update vectorial force */
 725             fix2             = _mm256_add_pd(fix2,tx);
 726             fiy2             = _mm256_add_pd(fiy2,ty);
 727             fiz2             = _mm256_add_pd(fiz2,tz);
 728
 729             fjx0             = _mm256_add_pd(fjx0,tx);
 730             fjy0             = _mm256_add_pd(fjy0,ty);
 731             fjz0             = _mm256_add_pd(fjz0,tz);
 732
 733             /**************************
 734              * CALCULATE INTERACTIONS *
 735              **************************/
 736
 737             /* Compute parameters for interactions between i and j atoms */
 738             qq30             = _mm256_mul_pd(iq3,jq0);
 739
 740             /* COULOMB ELECTROSTATICS */
 741             velec            = _mm256_mul_pd(qq30,rinv30);
 742             felec            = _mm256_mul_pd(velec,rinvsq30);
 743
 744             fscal            = felec;
 745
 746             /* Calculate temporary vectorial force */
 747             tx               = _mm256_mul_pd(fscal,dx30);
 748             ty               = _mm256_mul_pd(fscal,dy30);
 749             tz               = _mm256_mul_pd(fscal,dz30);
 750
 751             /* Update vectorial force */
 752             fix3             = _mm256_add_pd(fix3,tx);
 753             fiy3             = _mm256_add_pd(fiy3,ty);
 754             fiz3             = _mm256_add_pd(fiz3,tz);
 755
 756             fjx0             = _mm256_add_pd(fjx0,tx);
 757             fjy0             = _mm256_add_pd(fjy0,ty);
 758             fjz0             = _mm256_add_pd(fjz0,tz);
 759
 760             fjptrA             = f+j_coord_offsetA;
 761             fjptrB             = f+j_coord_offsetB;
 762             fjptrC             = f+j_coord_offsetC;
 763             fjptrD             = f+j_coord_offsetD;
 764
 765             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 766
 767             /* Inner loop uses 81 flops */
 768         }
 769
 770         if(jidx<j_index_end)
 771         {
 772
 773             /* Get j neighbor index, and coordinate index */
 774             jnrlistA         = jjnr[jidx];
 775             jnrlistB         = jjnr[jidx+1];
 776             jnrlistC         = jjnr[jidx+2];
 777             jnrlistD         = jjnr[jidx+3];
 778             /* Sign of each element will be negative for non-real atoms.
 779              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 780              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 781              */
 782             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 783
 784             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 785             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 786             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 787
 788             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 789             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 790             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 791             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 792             j_coord_offsetA  = DIM*jnrA;
 793             j_coord_offsetB  = DIM*jnrB;
 794             j_coord_offsetC  = DIM*jnrC;
 795             j_coord_offsetD  = DIM*jnrD;
 796
 797             /* load j atom coordinates */
 798             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 799                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 800                                                  &jx0,&jy0,&jz0);
 801
 802             /* Calculate displacement vector */
 803             dx10             = _mm256_sub_pd(ix1,jx0);
 804             dy10             = _mm256_sub_pd(iy1,jy0);
 805             dz10             = _mm256_sub_pd(iz1,jz0);
 806             dx20             = _mm256_sub_pd(ix2,jx0);
 807             dy20             = _mm256_sub_pd(iy2,jy0);
 808             dz20             = _mm256_sub_pd(iz2,jz0);
 809             dx30             = _mm256_sub_pd(ix3,jx0);
 810             dy30             = _mm256_sub_pd(iy3,jy0);
 811             dz30             = _mm256_sub_pd(iz3,jz0);
 812
 813             /* Calculate squared distance and things based on it */
 814             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 815             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 816             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 817
 818             rinv10           = avx256_invsqrt_d(rsq10);
 819             rinv20           = avx256_invsqrt_d(rsq20);
 820             rinv30           = avx256_invsqrt_d(rsq30);
 821
 822             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 823             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 824             rinvsq30         = _mm256_mul_pd(rinv30,rinv30);
 825
 826             /* Load parameters for j particles */
 827             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 828                                                                  charge+jnrC+0,charge+jnrD+0);
 829
 830             fjx0             = _mm256_setzero_pd();
 831             fjy0             = _mm256_setzero_pd();
 832             fjz0             = _mm256_setzero_pd();
 833
 834             /**************************
 835              * CALCULATE INTERACTIONS *
 836              **************************/
 837
 838             /* Compute parameters for interactions between i and j atoms */
 839             qq10             = _mm256_mul_pd(iq1,jq0);
 840
 841             /* COULOMB ELECTROSTATICS */
 842             velec            = _mm256_mul_pd(qq10,rinv10);
 843             felec            = _mm256_mul_pd(velec,rinvsq10);
 844
 845             fscal            = felec;
 846
 847             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 848
 849             /* Calculate temporary vectorial force */
 850             tx               = _mm256_mul_pd(fscal,dx10);
 851             ty               = _mm256_mul_pd(fscal,dy10);
 852             tz               = _mm256_mul_pd(fscal,dz10);
 853
 854             /* Update vectorial force */
 855             fix1             = _mm256_add_pd(fix1,tx);
 856             fiy1             = _mm256_add_pd(fiy1,ty);
 857             fiz1             = _mm256_add_pd(fiz1,tz);
 858
 859             fjx0             = _mm256_add_pd(fjx0,tx);
 860             fjy0             = _mm256_add_pd(fjy0,ty);
 861             fjz0             = _mm256_add_pd(fjz0,tz);
 862
 863             /**************************
 864              * CALCULATE INTERACTIONS *
 865              **************************/
 866
 867             /* Compute parameters for interactions between i and j atoms */
 868             qq20             = _mm256_mul_pd(iq2,jq0);
 869
 870             /* COULOMB ELECTROSTATICS */
 871             velec            = _mm256_mul_pd(qq20,rinv20);
 872             felec            = _mm256_mul_pd(velec,rinvsq20);
 873
 874             fscal            = felec;
 875
 876             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 877
 878             /* Calculate temporary vectorial force */
 879             tx               = _mm256_mul_pd(fscal,dx20);
 880             ty               = _mm256_mul_pd(fscal,dy20);
 881             tz               = _mm256_mul_pd(fscal,dz20);
 882
 883             /* Update vectorial force */
 884             fix2             = _mm256_add_pd(fix2,tx);
 885             fiy2             = _mm256_add_pd(fiy2,ty);
 886             fiz2             = _mm256_add_pd(fiz2,tz);
 887
 888             fjx0             = _mm256_add_pd(fjx0,tx);
 889             fjy0             = _mm256_add_pd(fjy0,ty);
 890             fjz0             = _mm256_add_pd(fjz0,tz);
 891
 892             /**************************
 893              * CALCULATE INTERACTIONS *
 894              **************************/
 895
 896             /* Compute parameters for interactions between i and j atoms */
 897             qq30             = _mm256_mul_pd(iq3,jq0);
 898
 899             /* COULOMB ELECTROSTATICS */
 900             velec            = _mm256_mul_pd(qq30,rinv30);
 901             felec            = _mm256_mul_pd(velec,rinvsq30);
 902
 903             fscal            = felec;
 904
 905             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 906
 907             /* Calculate temporary vectorial force */
 908             tx               = _mm256_mul_pd(fscal,dx30);
 909             ty               = _mm256_mul_pd(fscal,dy30);
 910             tz               = _mm256_mul_pd(fscal,dz30);
 911
 912             /* Update vectorial force */
 913             fix3             = _mm256_add_pd(fix3,tx);
 914             fiy3             = _mm256_add_pd(fiy3,ty);
 915             fiz3             = _mm256_add_pd(fiz3,tz);
 916
 917             fjx0             = _mm256_add_pd(fjx0,tx);
 918             fjy0             = _mm256_add_pd(fjy0,ty);
 919             fjz0             = _mm256_add_pd(fjz0,tz);
 920
 921             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 922             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 923             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 924             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 925
 926             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 927
 928             /* Inner loop uses 81 flops */
 929         }
 930
 931         /* End of innermost loop */
 932
 933         gmx_mm256_update_iforce_3atom_swizzle_pd(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 934                                                  f+i_coord_offset+DIM,fshift+i_shift_offset);
 935
 936         /* Increment number of inner iterations */
 937         inneriter                  += j_index_end - j_index_start;
 938
 939         /* Outer loop uses 18 flops */
 940     }
 941
 942     /* Increment number of outer iterations */
 943     outeriter        += nri;
 944
 945     /* Update outer/inner flops */
 946
 947     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4_F,outeriter*18 + inneriter*81);
 948 }