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, 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 "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_256_double.h"
  48 #include "kernelutil_x86_avx_256_double.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwNone_GeomW4P1_VF_avx_256_double
  52  * Electrostatics interaction: Coulomb
  53  * VdW interaction:            None
  54  * Geometry:                   Water4-Particle
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecCoul_VdwNone_GeomW4P1_VF_avx_256_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,C,D refer to j loop unrolling done with AVX, e.g. for the four 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,jnrC,jnrD;
  75     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  76     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  77     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  78     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  79     real             rcutoff_scalar;
  80     real             *shiftvec,*fshift,*x,*f;
  81     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  82     real             scratch[4*DIM];
  83     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  84     real *           vdwioffsetptr1;
  85     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  86     real *           vdwioffsetptr2;
  87     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  88     real *           vdwioffsetptr3;
  89     __m256d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  90     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  91     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  92     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  93     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  94     __m256d          dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
  95     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
  96     real             *charge;
  97     __m256d          dummy_mask,cutoff_mask;
  98     __m128           tmpmask0,tmpmask1;
  99     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 100     __m256d          one     = _mm256_set1_pd(1.0);
 101     __m256d          two     = _mm256_set1_pd(2.0);
 102     x                = xx[0];
 103     f                = ff[0];
 104
 105     nri              = nlist->nri;
 106     iinr             = nlist->iinr;
 107     jindex           = nlist->jindex;
 108     jjnr             = nlist->jjnr;
 109     shiftidx         = nlist->shift;
 110     gid              = nlist->gid;
 111     shiftvec         = fr->shift_vec[0];
 112     fshift           = fr->fshift[0];
 113     facel            = _mm256_set1_pd(fr->epsfac);
 114     charge           = mdatoms->chargeA;
 115
 116     /* Setup water-specific parameters */
 117     inr              = nlist->iinr[0];
 118     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 119     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 120     iq3              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
 121
 122     /* Avoid stupid compiler warnings */
 123     jnrA = jnrB = jnrC = jnrD = 0;
 124     j_coord_offsetA = 0;
 125     j_coord_offsetB = 0;
 126     j_coord_offsetC = 0;
 127     j_coord_offsetD = 0;
 128
 129     outeriter        = 0;
 130     inneriter        = 0;
 131
 132     for(iidx=0;iidx<4*DIM;iidx++)
 133     {
 134         scratch[iidx] = 0.0;
 135     }
 136
 137     /* Start outer loop over neighborlists */
 138     for(iidx=0; iidx<nri; iidx++)
 139     {
 140         /* Load shift vector for this list */
 141         i_shift_offset   = DIM*shiftidx[iidx];
 142
 143         /* Load limits for loop over neighbors */
 144         j_index_start    = jindex[iidx];
 145         j_index_end      = jindex[iidx+1];
 146
 147         /* Get outer coordinate index */
 148         inr              = iinr[iidx];
 149         i_coord_offset   = DIM*inr;
 150
 151         /* Load i particle coords and add shift vector */
 152         gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset+DIM,
 153                                                     &ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 154
 155         fix1             = _mm256_setzero_pd();
 156         fiy1             = _mm256_setzero_pd();
 157         fiz1             = _mm256_setzero_pd();
 158         fix2             = _mm256_setzero_pd();
 159         fiy2             = _mm256_setzero_pd();
 160         fiz2             = _mm256_setzero_pd();
 161         fix3             = _mm256_setzero_pd();
 162         fiy3             = _mm256_setzero_pd();
 163         fiz3             = _mm256_setzero_pd();
 164
 165         /* Reset potential sums */
 166         velecsum         = _mm256_setzero_pd();
 167
 168         /* Start inner kernel loop */
 169         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 170         {
 171
 172             /* Get j neighbor index, and coordinate index */
 173             jnrA             = jjnr[jidx];
 174             jnrB             = jjnr[jidx+1];
 175             jnrC             = jjnr[jidx+2];
 176             jnrD             = jjnr[jidx+3];
 177             j_coord_offsetA  = DIM*jnrA;
 178             j_coord_offsetB  = DIM*jnrB;
 179             j_coord_offsetC  = DIM*jnrC;
 180             j_coord_offsetD  = DIM*jnrD;
 181
 182             /* load j atom coordinates */
 183             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 184                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 185                                                  &jx0,&jy0,&jz0);
 186
 187             /* Calculate displacement vector */
 188             dx10             = _mm256_sub_pd(ix1,jx0);
 189             dy10             = _mm256_sub_pd(iy1,jy0);
 190             dz10             = _mm256_sub_pd(iz1,jz0);
 191             dx20             = _mm256_sub_pd(ix2,jx0);
 192             dy20             = _mm256_sub_pd(iy2,jy0);
 193             dz20             = _mm256_sub_pd(iz2,jz0);
 194             dx30             = _mm256_sub_pd(ix3,jx0);
 195             dy30             = _mm256_sub_pd(iy3,jy0);
 196             dz30             = _mm256_sub_pd(iz3,jz0);
 197
 198             /* Calculate squared distance and things based on it */
 199             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 200             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 201             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 202
 203             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 204             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 205             rinv30           = gmx_mm256_invsqrt_pd(rsq30);
 206
 207             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 208             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 209             rinvsq30         = _mm256_mul_pd(rinv30,rinv30);
 210
 211             /* Load parameters for j particles */
 212             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 213                                                                  charge+jnrC+0,charge+jnrD+0);
 214
 215             fjx0             = _mm256_setzero_pd();
 216             fjy0             = _mm256_setzero_pd();
 217             fjz0             = _mm256_setzero_pd();
 218
 219             /**************************
 220              * CALCULATE INTERACTIONS *
 221              **************************/
 222
 223             /* Compute parameters for interactions between i and j atoms */
 224             qq10             = _mm256_mul_pd(iq1,jq0);
 225
 226             /* COULOMB ELECTROSTATICS */
 227             velec            = _mm256_mul_pd(qq10,rinv10);
 228             felec            = _mm256_mul_pd(velec,rinvsq10);
 229
 230             /* Update potential sum for this i atom from the interaction with this j atom. */
 231             velecsum         = _mm256_add_pd(velecsum,velec);
 232
 233             fscal            = felec;
 234
 235             /* Calculate temporary vectorial force */
 236             tx               = _mm256_mul_pd(fscal,dx10);
 237             ty               = _mm256_mul_pd(fscal,dy10);
 238             tz               = _mm256_mul_pd(fscal,dz10);
 239
 240             /* Update vectorial force */
 241             fix1             = _mm256_add_pd(fix1,tx);
 242             fiy1             = _mm256_add_pd(fiy1,ty);
 243             fiz1             = _mm256_add_pd(fiz1,tz);
 244
 245             fjx0             = _mm256_add_pd(fjx0,tx);
 246             fjy0             = _mm256_add_pd(fjy0,ty);
 247             fjz0             = _mm256_add_pd(fjz0,tz);
 248
 249             /**************************
 250              * CALCULATE INTERACTIONS *
 251              **************************/
 252
 253             /* Compute parameters for interactions between i and j atoms */
 254             qq20             = _mm256_mul_pd(iq2,jq0);
 255
 256             /* COULOMB ELECTROSTATICS */
 257             velec            = _mm256_mul_pd(qq20,rinv20);
 258             felec            = _mm256_mul_pd(velec,rinvsq20);
 259
 260             /* Update potential sum for this i atom from the interaction with this j atom. */
 261             velecsum         = _mm256_add_pd(velecsum,velec);
 262
 263             fscal            = felec;
 264
 265             /* Calculate temporary vectorial force */
 266             tx               = _mm256_mul_pd(fscal,dx20);
 267             ty               = _mm256_mul_pd(fscal,dy20);
 268             tz               = _mm256_mul_pd(fscal,dz20);
 269
 270             /* Update vectorial force */
 271             fix2             = _mm256_add_pd(fix2,tx);
 272             fiy2             = _mm256_add_pd(fiy2,ty);
 273             fiz2             = _mm256_add_pd(fiz2,tz);
 274
 275             fjx0             = _mm256_add_pd(fjx0,tx);
 276             fjy0             = _mm256_add_pd(fjy0,ty);
 277             fjz0             = _mm256_add_pd(fjz0,tz);
 278
 279             /**************************
 280              * CALCULATE INTERACTIONS *
 281              **************************/
 282
 283             /* Compute parameters for interactions between i and j atoms */
 284             qq30             = _mm256_mul_pd(iq3,jq0);
 285
 286             /* COULOMB ELECTROSTATICS */
 287             velec            = _mm256_mul_pd(qq30,rinv30);
 288             felec            = _mm256_mul_pd(velec,rinvsq30);
 289
 290             /* Update potential sum for this i atom from the interaction with this j atom. */
 291             velecsum         = _mm256_add_pd(velecsum,velec);
 292
 293             fscal            = felec;
 294
 295             /* Calculate temporary vectorial force */
 296             tx               = _mm256_mul_pd(fscal,dx30);
 297             ty               = _mm256_mul_pd(fscal,dy30);
 298             tz               = _mm256_mul_pd(fscal,dz30);
 299
 300             /* Update vectorial force */
 301             fix3             = _mm256_add_pd(fix3,tx);
 302             fiy3             = _mm256_add_pd(fiy3,ty);
 303             fiz3             = _mm256_add_pd(fiz3,tz);
 304
 305             fjx0             = _mm256_add_pd(fjx0,tx);
 306             fjy0             = _mm256_add_pd(fjy0,ty);
 307             fjz0             = _mm256_add_pd(fjz0,tz);
 308
 309             fjptrA             = f+j_coord_offsetA;
 310             fjptrB             = f+j_coord_offsetB;
 311             fjptrC             = f+j_coord_offsetC;
 312             fjptrD             = f+j_coord_offsetD;
 313
 314             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 315
 316             /* Inner loop uses 84 flops */
 317         }
 318
 319         if(jidx<j_index_end)
 320         {
 321
 322             /* Get j neighbor index, and coordinate index */
 323             jnrlistA         = jjnr[jidx];
 324             jnrlistB         = jjnr[jidx+1];
 325             jnrlistC         = jjnr[jidx+2];
 326             jnrlistD         = jjnr[jidx+3];
 327             /* Sign of each element will be negative for non-real atoms.
 328              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 329              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 330              */
 331             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 332
 333             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 334             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 335             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 336
 337             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 338             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 339             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 340             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 341             j_coord_offsetA  = DIM*jnrA;
 342             j_coord_offsetB  = DIM*jnrB;
 343             j_coord_offsetC  = DIM*jnrC;
 344             j_coord_offsetD  = DIM*jnrD;
 345
 346             /* load j atom coordinates */
 347             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 348                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 349                                                  &jx0,&jy0,&jz0);
 350
 351             /* Calculate displacement vector */
 352             dx10             = _mm256_sub_pd(ix1,jx0);
 353             dy10             = _mm256_sub_pd(iy1,jy0);
 354             dz10             = _mm256_sub_pd(iz1,jz0);
 355             dx20             = _mm256_sub_pd(ix2,jx0);
 356             dy20             = _mm256_sub_pd(iy2,jy0);
 357             dz20             = _mm256_sub_pd(iz2,jz0);
 358             dx30             = _mm256_sub_pd(ix3,jx0);
 359             dy30             = _mm256_sub_pd(iy3,jy0);
 360             dz30             = _mm256_sub_pd(iz3,jz0);
 361
 362             /* Calculate squared distance and things based on it */
 363             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 364             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 365             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 366
 367             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 368             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 369             rinv30           = gmx_mm256_invsqrt_pd(rsq30);
 370
 371             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 372             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 373             rinvsq30         = _mm256_mul_pd(rinv30,rinv30);
 374
 375             /* Load parameters for j particles */
 376             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 377                                                                  charge+jnrC+0,charge+jnrD+0);
 378
 379             fjx0             = _mm256_setzero_pd();
 380             fjy0             = _mm256_setzero_pd();
 381             fjz0             = _mm256_setzero_pd();
 382
 383             /**************************
 384              * CALCULATE INTERACTIONS *
 385              **************************/
 386
 387             /* Compute parameters for interactions between i and j atoms */
 388             qq10             = _mm256_mul_pd(iq1,jq0);
 389
 390             /* COULOMB ELECTROSTATICS */
 391             velec            = _mm256_mul_pd(qq10,rinv10);
 392             felec            = _mm256_mul_pd(velec,rinvsq10);
 393
 394             /* Update potential sum for this i atom from the interaction with this j atom. */
 395             velec            = _mm256_andnot_pd(dummy_mask,velec);
 396             velecsum         = _mm256_add_pd(velecsum,velec);
 397
 398             fscal            = felec;
 399
 400             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 401
 402             /* Calculate temporary vectorial force */
 403             tx               = _mm256_mul_pd(fscal,dx10);
 404             ty               = _mm256_mul_pd(fscal,dy10);
 405             tz               = _mm256_mul_pd(fscal,dz10);
 406
 407             /* Update vectorial force */
 408             fix1             = _mm256_add_pd(fix1,tx);
 409             fiy1             = _mm256_add_pd(fiy1,ty);
 410             fiz1             = _mm256_add_pd(fiz1,tz);
 411
 412             fjx0             = _mm256_add_pd(fjx0,tx);
 413             fjy0             = _mm256_add_pd(fjy0,ty);
 414             fjz0             = _mm256_add_pd(fjz0,tz);
 415
 416             /**************************
 417              * CALCULATE INTERACTIONS *
 418              **************************/
 419
 420             /* Compute parameters for interactions between i and j atoms */
 421             qq20             = _mm256_mul_pd(iq2,jq0);
 422
 423             /* COULOMB ELECTROSTATICS */
 424             velec            = _mm256_mul_pd(qq20,rinv20);
 425             felec            = _mm256_mul_pd(velec,rinvsq20);
 426
 427             /* Update potential sum for this i atom from the interaction with this j atom. */
 428             velec            = _mm256_andnot_pd(dummy_mask,velec);
 429             velecsum         = _mm256_add_pd(velecsum,velec);
 430
 431             fscal            = felec;
 432
 433             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 434
 435             /* Calculate temporary vectorial force */
 436             tx               = _mm256_mul_pd(fscal,dx20);
 437             ty               = _mm256_mul_pd(fscal,dy20);
 438             tz               = _mm256_mul_pd(fscal,dz20);
 439
 440             /* Update vectorial force */
 441             fix2             = _mm256_add_pd(fix2,tx);
 442             fiy2             = _mm256_add_pd(fiy2,ty);
 443             fiz2             = _mm256_add_pd(fiz2,tz);
 444
 445             fjx0             = _mm256_add_pd(fjx0,tx);
 446             fjy0             = _mm256_add_pd(fjy0,ty);
 447             fjz0             = _mm256_add_pd(fjz0,tz);
 448
 449             /**************************
 450              * CALCULATE INTERACTIONS *
 451              **************************/
 452
 453             /* Compute parameters for interactions between i and j atoms */
 454             qq30             = _mm256_mul_pd(iq3,jq0);
 455
 456             /* COULOMB ELECTROSTATICS */
 457             velec            = _mm256_mul_pd(qq30,rinv30);
 458             felec            = _mm256_mul_pd(velec,rinvsq30);
 459
 460             /* Update potential sum for this i atom from the interaction with this j atom. */
 461             velec            = _mm256_andnot_pd(dummy_mask,velec);
 462             velecsum         = _mm256_add_pd(velecsum,velec);
 463
 464             fscal            = felec;
 465
 466             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 467
 468             /* Calculate temporary vectorial force */
 469             tx               = _mm256_mul_pd(fscal,dx30);
 470             ty               = _mm256_mul_pd(fscal,dy30);
 471             tz               = _mm256_mul_pd(fscal,dz30);
 472
 473             /* Update vectorial force */
 474             fix3             = _mm256_add_pd(fix3,tx);
 475             fiy3             = _mm256_add_pd(fiy3,ty);
 476             fiz3             = _mm256_add_pd(fiz3,tz);
 477
 478             fjx0             = _mm256_add_pd(fjx0,tx);
 479             fjy0             = _mm256_add_pd(fjy0,ty);
 480             fjz0             = _mm256_add_pd(fjz0,tz);
 481
 482             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 483             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 484             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 485             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 486
 487             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 488
 489             /* Inner loop uses 84 flops */
 490         }
 491
 492         /* End of innermost loop */
 493
 494         gmx_mm256_update_iforce_3atom_swizzle_pd(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 495                                                  f+i_coord_offset+DIM,fshift+i_shift_offset);
 496
 497         ggid                        = gid[iidx];
 498         /* Update potential energies */
 499         gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
 500
 501         /* Increment number of inner iterations */
 502         inneriter                  += j_index_end - j_index_start;
 503
 504         /* Outer loop uses 19 flops */
 505     }
 506
 507     /* Increment number of outer iterations */
 508     outeriter        += nri;
 509
 510     /* Update outer/inner flops */
 511
 512     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4_VF,outeriter*19 + inneriter*84);
 513 }
 514 /*
 515  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwNone_GeomW4P1_F_avx_256_double
 516  * Electrostatics interaction: Coulomb
 517  * VdW interaction:            None
 518  * Geometry:                   Water4-Particle
 519  * Calculate force/pot:        Force
 520  */
 521 void
 522 nb_kernel_ElecCoul_VdwNone_GeomW4P1_F_avx_256_double
 523                     (t_nblist                    * gmx_restrict       nlist,
 524                      rvec                        * gmx_restrict          xx,
 525                      rvec                        * gmx_restrict          ff,
 526                      t_forcerec                  * gmx_restrict          fr,
 527                      t_mdatoms                   * gmx_restrict     mdatoms,
 528                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 529                      t_nrnb                      * gmx_restrict        nrnb)
 530 {
 531     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 532      * just 0 for non-waters.
 533      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
 534      * jnr indices corresponding to data put in the four positions in the SIMD register.
 535      */
 536     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 537     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 538     int              jnrA,jnrB,jnrC,jnrD;
 539     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 540     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 541     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 542     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 543     real             rcutoff_scalar;
 544     real             *shiftvec,*fshift,*x,*f;
 545     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 546     real             scratch[4*DIM];
 547     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 548     real *           vdwioffsetptr1;
 549     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 550     real *           vdwioffsetptr2;
 551     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 552     real *           vdwioffsetptr3;
 553     __m256d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
 554     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 555     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 556     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 557     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 558     __m256d          dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
 559     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
 560     real             *charge;
 561     __m256d          dummy_mask,cutoff_mask;
 562     __m128           tmpmask0,tmpmask1;
 563     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 564     __m256d          one     = _mm256_set1_pd(1.0);
 565     __m256d          two     = _mm256_set1_pd(2.0);
 566     x                = xx[0];
 567     f                = ff[0];
 568
 569     nri              = nlist->nri;
 570     iinr             = nlist->iinr;
 571     jindex           = nlist->jindex;
 572     jjnr             = nlist->jjnr;
 573     shiftidx         = nlist->shift;
 574     gid              = nlist->gid;
 575     shiftvec         = fr->shift_vec[0];
 576     fshift           = fr->fshift[0];
 577     facel            = _mm256_set1_pd(fr->epsfac);
 578     charge           = mdatoms->chargeA;
 579
 580     /* Setup water-specific parameters */
 581     inr              = nlist->iinr[0];
 582     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 583     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 584     iq3              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
 585
 586     /* Avoid stupid compiler warnings */
 587     jnrA = jnrB = jnrC = jnrD = 0;
 588     j_coord_offsetA = 0;
 589     j_coord_offsetB = 0;
 590     j_coord_offsetC = 0;
 591     j_coord_offsetD = 0;
 592
 593     outeriter        = 0;
 594     inneriter        = 0;
 595
 596     for(iidx=0;iidx<4*DIM;iidx++)
 597     {
 598         scratch[iidx] = 0.0;
 599     }
 600
 601     /* Start outer loop over neighborlists */
 602     for(iidx=0; iidx<nri; iidx++)
 603     {
 604         /* Load shift vector for this list */
 605         i_shift_offset   = DIM*shiftidx[iidx];
 606
 607         /* Load limits for loop over neighbors */
 608         j_index_start    = jindex[iidx];
 609         j_index_end      = jindex[iidx+1];
 610
 611         /* Get outer coordinate index */
 612         inr              = iinr[iidx];
 613         i_coord_offset   = DIM*inr;
 614
 615         /* Load i particle coords and add shift vector */
 616         gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset+DIM,
 617                                                     &ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 618
 619         fix1             = _mm256_setzero_pd();
 620         fiy1             = _mm256_setzero_pd();
 621         fiz1             = _mm256_setzero_pd();
 622         fix2             = _mm256_setzero_pd();
 623         fiy2             = _mm256_setzero_pd();
 624         fiz2             = _mm256_setzero_pd();
 625         fix3             = _mm256_setzero_pd();
 626         fiy3             = _mm256_setzero_pd();
 627         fiz3             = _mm256_setzero_pd();
 628
 629         /* Start inner kernel loop */
 630         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 631         {
 632
 633             /* Get j neighbor index, and coordinate index */
 634             jnrA             = jjnr[jidx];
 635             jnrB             = jjnr[jidx+1];
 636             jnrC             = jjnr[jidx+2];
 637             jnrD             = jjnr[jidx+3];
 638             j_coord_offsetA  = DIM*jnrA;
 639             j_coord_offsetB  = DIM*jnrB;
 640             j_coord_offsetC  = DIM*jnrC;
 641             j_coord_offsetD  = DIM*jnrD;
 642
 643             /* load j atom coordinates */
 644             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 645                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 646                                                  &jx0,&jy0,&jz0);
 647
 648             /* Calculate displacement vector */
 649             dx10             = _mm256_sub_pd(ix1,jx0);
 650             dy10             = _mm256_sub_pd(iy1,jy0);
 651             dz10             = _mm256_sub_pd(iz1,jz0);
 652             dx20             = _mm256_sub_pd(ix2,jx0);
 653             dy20             = _mm256_sub_pd(iy2,jy0);
 654             dz20             = _mm256_sub_pd(iz2,jz0);
 655             dx30             = _mm256_sub_pd(ix3,jx0);
 656             dy30             = _mm256_sub_pd(iy3,jy0);
 657             dz30             = _mm256_sub_pd(iz3,jz0);
 658
 659             /* Calculate squared distance and things based on it */
 660             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 661             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 662             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 663
 664             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 665             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 666             rinv30           = gmx_mm256_invsqrt_pd(rsq30);
 667
 668             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 669             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 670             rinvsq30         = _mm256_mul_pd(rinv30,rinv30);
 671
 672             /* Load parameters for j particles */
 673             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 674                                                                  charge+jnrC+0,charge+jnrD+0);
 675
 676             fjx0             = _mm256_setzero_pd();
 677             fjy0             = _mm256_setzero_pd();
 678             fjz0             = _mm256_setzero_pd();
 679
 680             /**************************
 681              * CALCULATE INTERACTIONS *
 682              **************************/
 683
 684             /* Compute parameters for interactions between i and j atoms */
 685             qq10             = _mm256_mul_pd(iq1,jq0);
 686
 687             /* COULOMB ELECTROSTATICS */
 688             velec            = _mm256_mul_pd(qq10,rinv10);
 689             felec            = _mm256_mul_pd(velec,rinvsq10);
 690
 691             fscal            = felec;
 692
 693             /* Calculate temporary vectorial force */
 694             tx               = _mm256_mul_pd(fscal,dx10);
 695             ty               = _mm256_mul_pd(fscal,dy10);
 696             tz               = _mm256_mul_pd(fscal,dz10);
 697
 698             /* Update vectorial force */
 699             fix1             = _mm256_add_pd(fix1,tx);
 700             fiy1             = _mm256_add_pd(fiy1,ty);
 701             fiz1             = _mm256_add_pd(fiz1,tz);
 702
 703             fjx0             = _mm256_add_pd(fjx0,tx);
 704             fjy0             = _mm256_add_pd(fjy0,ty);
 705             fjz0             = _mm256_add_pd(fjz0,tz);
 706
 707             /**************************
 708              * CALCULATE INTERACTIONS *
 709              **************************/
 710
 711             /* Compute parameters for interactions between i and j atoms */
 712             qq20             = _mm256_mul_pd(iq2,jq0);
 713
 714             /* COULOMB ELECTROSTATICS */
 715             velec            = _mm256_mul_pd(qq20,rinv20);
 716             felec            = _mm256_mul_pd(velec,rinvsq20);
 717
 718             fscal            = felec;
 719
 720             /* Calculate temporary vectorial force */
 721             tx               = _mm256_mul_pd(fscal,dx20);
 722             ty               = _mm256_mul_pd(fscal,dy20);
 723             tz               = _mm256_mul_pd(fscal,dz20);
 724
 725             /* Update vectorial force */
 726             fix2             = _mm256_add_pd(fix2,tx);
 727             fiy2             = _mm256_add_pd(fiy2,ty);
 728             fiz2             = _mm256_add_pd(fiz2,tz);
 729
 730             fjx0             = _mm256_add_pd(fjx0,tx);
 731             fjy0             = _mm256_add_pd(fjy0,ty);
 732             fjz0             = _mm256_add_pd(fjz0,tz);
 733
 734             /**************************
 735              * CALCULATE INTERACTIONS *
 736              **************************/
 737
 738             /* Compute parameters for interactions between i and j atoms */
 739             qq30             = _mm256_mul_pd(iq3,jq0);
 740
 741             /* COULOMB ELECTROSTATICS */
 742             velec            = _mm256_mul_pd(qq30,rinv30);
 743             felec            = _mm256_mul_pd(velec,rinvsq30);
 744
 745             fscal            = felec;
 746
 747             /* Calculate temporary vectorial force */
 748             tx               = _mm256_mul_pd(fscal,dx30);
 749             ty               = _mm256_mul_pd(fscal,dy30);
 750             tz               = _mm256_mul_pd(fscal,dz30);
 751
 752             /* Update vectorial force */
 753             fix3             = _mm256_add_pd(fix3,tx);
 754             fiy3             = _mm256_add_pd(fiy3,ty);
 755             fiz3             = _mm256_add_pd(fiz3,tz);
 756
 757             fjx0             = _mm256_add_pd(fjx0,tx);
 758             fjy0             = _mm256_add_pd(fjy0,ty);
 759             fjz0             = _mm256_add_pd(fjz0,tz);
 760
 761             fjptrA             = f+j_coord_offsetA;
 762             fjptrB             = f+j_coord_offsetB;
 763             fjptrC             = f+j_coord_offsetC;
 764             fjptrD             = f+j_coord_offsetD;
 765
 766             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 767
 768             /* Inner loop uses 81 flops */
 769         }
 770
 771         if(jidx<j_index_end)
 772         {
 773
 774             /* Get j neighbor index, and coordinate index */
 775             jnrlistA         = jjnr[jidx];
 776             jnrlistB         = jjnr[jidx+1];
 777             jnrlistC         = jjnr[jidx+2];
 778             jnrlistD         = jjnr[jidx+3];
 779             /* Sign of each element will be negative for non-real atoms.
 780              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 781              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 782              */
 783             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 784
 785             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 786             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 787             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 788
 789             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 790             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 791             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 792             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 793             j_coord_offsetA  = DIM*jnrA;
 794             j_coord_offsetB  = DIM*jnrB;
 795             j_coord_offsetC  = DIM*jnrC;
 796             j_coord_offsetD  = DIM*jnrD;
 797
 798             /* load j atom coordinates */
 799             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 800                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 801                                                  &jx0,&jy0,&jz0);
 802
 803             /* Calculate displacement vector */
 804             dx10             = _mm256_sub_pd(ix1,jx0);
 805             dy10             = _mm256_sub_pd(iy1,jy0);
 806             dz10             = _mm256_sub_pd(iz1,jz0);
 807             dx20             = _mm256_sub_pd(ix2,jx0);
 808             dy20             = _mm256_sub_pd(iy2,jy0);
 809             dz20             = _mm256_sub_pd(iz2,jz0);
 810             dx30             = _mm256_sub_pd(ix3,jx0);
 811             dy30             = _mm256_sub_pd(iy3,jy0);
 812             dz30             = _mm256_sub_pd(iz3,jz0);
 813
 814             /* Calculate squared distance and things based on it */
 815             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 816             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 817             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 818
 819             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 820             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 821             rinv30           = gmx_mm256_invsqrt_pd(rsq30);
 822
 823             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 824             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 825             rinvsq30         = _mm256_mul_pd(rinv30,rinv30);
 826
 827             /* Load parameters for j particles */
 828             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 829                                                                  charge+jnrC+0,charge+jnrD+0);
 830
 831             fjx0             = _mm256_setzero_pd();
 832             fjy0             = _mm256_setzero_pd();
 833             fjz0             = _mm256_setzero_pd();
 834
 835             /**************************
 836              * CALCULATE INTERACTIONS *
 837              **************************/
 838
 839             /* Compute parameters for interactions between i and j atoms */
 840             qq10             = _mm256_mul_pd(iq1,jq0);
 841
 842             /* COULOMB ELECTROSTATICS */
 843             velec            = _mm256_mul_pd(qq10,rinv10);
 844             felec            = _mm256_mul_pd(velec,rinvsq10);
 845
 846             fscal            = felec;
 847
 848             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 849
 850             /* Calculate temporary vectorial force */
 851             tx               = _mm256_mul_pd(fscal,dx10);
 852             ty               = _mm256_mul_pd(fscal,dy10);
 853             tz               = _mm256_mul_pd(fscal,dz10);
 854
 855             /* Update vectorial force */
 856             fix1             = _mm256_add_pd(fix1,tx);
 857             fiy1             = _mm256_add_pd(fiy1,ty);
 858             fiz1             = _mm256_add_pd(fiz1,tz);
 859
 860             fjx0             = _mm256_add_pd(fjx0,tx);
 861             fjy0             = _mm256_add_pd(fjy0,ty);
 862             fjz0             = _mm256_add_pd(fjz0,tz);
 863
 864             /**************************
 865              * CALCULATE INTERACTIONS *
 866              **************************/
 867
 868             /* Compute parameters for interactions between i and j atoms */
 869             qq20             = _mm256_mul_pd(iq2,jq0);
 870
 871             /* COULOMB ELECTROSTATICS */
 872             velec            = _mm256_mul_pd(qq20,rinv20);
 873             felec            = _mm256_mul_pd(velec,rinvsq20);
 874
 875             fscal            = felec;
 876
 877             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 878
 879             /* Calculate temporary vectorial force */
 880             tx               = _mm256_mul_pd(fscal,dx20);
 881             ty               = _mm256_mul_pd(fscal,dy20);
 882             tz               = _mm256_mul_pd(fscal,dz20);
 883
 884             /* Update vectorial force */
 885             fix2             = _mm256_add_pd(fix2,tx);
 886             fiy2             = _mm256_add_pd(fiy2,ty);
 887             fiz2             = _mm256_add_pd(fiz2,tz);
 888
 889             fjx0             = _mm256_add_pd(fjx0,tx);
 890             fjy0             = _mm256_add_pd(fjy0,ty);
 891             fjz0             = _mm256_add_pd(fjz0,tz);
 892
 893             /**************************
 894              * CALCULATE INTERACTIONS *
 895              **************************/
 896
 897             /* Compute parameters for interactions between i and j atoms */
 898             qq30             = _mm256_mul_pd(iq3,jq0);
 899
 900             /* COULOMB ELECTROSTATICS */
 901             velec            = _mm256_mul_pd(qq30,rinv30);
 902             felec            = _mm256_mul_pd(velec,rinvsq30);
 903
 904             fscal            = felec;
 905
 906             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 907
 908             /* Calculate temporary vectorial force */
 909             tx               = _mm256_mul_pd(fscal,dx30);
 910             ty               = _mm256_mul_pd(fscal,dy30);
 911             tz               = _mm256_mul_pd(fscal,dz30);
 912
 913             /* Update vectorial force */
 914             fix3             = _mm256_add_pd(fix3,tx);
 915             fiy3             = _mm256_add_pd(fiy3,ty);
 916             fiz3             = _mm256_add_pd(fiz3,tz);
 917
 918             fjx0             = _mm256_add_pd(fjx0,tx);
 919             fjy0             = _mm256_add_pd(fjy0,ty);
 920             fjz0             = _mm256_add_pd(fjz0,tz);
 921
 922             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 923             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 924             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 925             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 926
 927             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 928
 929             /* Inner loop uses 81 flops */
 930         }
 931
 932         /* End of innermost loop */
 933
 934         gmx_mm256_update_iforce_3atom_swizzle_pd(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 935                                                  f+i_coord_offset+DIM,fshift+i_shift_offset);
 936
 937         /* Increment number of inner iterations */
 938         inneriter                  += j_index_end - j_index_start;
 939
 940         /* Outer loop uses 18 flops */
 941     }
 942
 943     /* Increment number of outer iterations */
 944     outeriter        += nri;
 945
 946     /* Update outer/inner flops */
 947
 948     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4_F,outeriter*18 + inneriter*81);
 949 }