src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_single/nb_kernel_ElecNone_VdwLJ_GeomP1P1_avx_256_single.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_single 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/legacyheaders/types/simple.h"
  46 #include "gromacs/math/vec.h"
  47 #include "gromacs/legacyheaders/nrnb.h"
  48
  49 #include "gromacs/simd/math_x86_avx_256_single.h"
  50 #include "kernelutil_x86_avx_256_single.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecNone_VdwLJ_GeomP1P1_VF_avx_256_single
  54  * Electrostatics interaction: None
  55  * VdW interaction:            LennardJones
  56  * Geometry:                   Particle-Particle
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecNone_VdwLJ_GeomP1P1_VF_avx_256_single
  61                     (t_nblist                    * gmx_restrict       nlist,
  62                      rvec                        * gmx_restrict          xx,
  63                      rvec                        * gmx_restrict          ff,
  64                      t_forcerec                  * gmx_restrict          fr,
  65                      t_mdatoms                   * gmx_restrict     mdatoms,
  66                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  67                      t_nrnb                      * gmx_restrict        nrnb)
  68 {
  69     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  70      * just 0 for non-waters.
  71      * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
  72      * jnr indices corresponding to data put in the four positions in the SIMD register.
  73      */
  74     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  75     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  76     int              jnrA,jnrB,jnrC,jnrD;
  77     int              jnrE,jnrF,jnrG,jnrH;
  78     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  79     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  80     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  81     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
  82     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  83     real             rcutoff_scalar;
  84     real             *shiftvec,*fshift,*x,*f;
  85     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
  86     real             scratch[4*DIM];
  87     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  88     real *           vdwioffsetptr0;
  89     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  90     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
  91     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  92     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  93     int              nvdwtype;
  94     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  95     int              *vdwtype;
  96     real             *vdwparam;
  97     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
  98     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
  99     __m256           dummy_mask,cutoff_mask;
 100     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 101     __m256           one     = _mm256_set1_ps(1.0);
 102     __m256           two     = _mm256_set1_ps(2.0);
 103     x                = xx[0];
 104     f                = ff[0];
 105
 106     nri              = nlist->nri;
 107     iinr             = nlist->iinr;
 108     jindex           = nlist->jindex;
 109     jjnr             = nlist->jjnr;
 110     shiftidx         = nlist->shift;
 111     gid              = nlist->gid;
 112     shiftvec         = fr->shift_vec[0];
 113     fshift           = fr->fshift[0];
 114     nvdwtype         = fr->ntype;
 115     vdwparam         = fr->nbfp;
 116     vdwtype          = mdatoms->typeA;
 117
 118     /* Avoid stupid compiler warnings */
 119     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 120     j_coord_offsetA = 0;
 121     j_coord_offsetB = 0;
 122     j_coord_offsetC = 0;
 123     j_coord_offsetD = 0;
 124     j_coord_offsetE = 0;
 125     j_coord_offsetF = 0;
 126     j_coord_offsetG = 0;
 127     j_coord_offsetH = 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_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 153
 154         fix0             = _mm256_setzero_ps();
 155         fiy0             = _mm256_setzero_ps();
 156         fiz0             = _mm256_setzero_ps();
 157
 158         /* Load parameters for i particles */
 159         vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 160
 161         /* Reset potential sums */
 162         vvdwsum          = _mm256_setzero_ps();
 163
 164         /* Start inner kernel loop */
 165         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 166         {
 167
 168             /* Get j neighbor index, and coordinate index */
 169             jnrA             = jjnr[jidx];
 170             jnrB             = jjnr[jidx+1];
 171             jnrC             = jjnr[jidx+2];
 172             jnrD             = jjnr[jidx+3];
 173             jnrE             = jjnr[jidx+4];
 174             jnrF             = jjnr[jidx+5];
 175             jnrG             = jjnr[jidx+6];
 176             jnrH             = jjnr[jidx+7];
 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             j_coord_offsetE  = DIM*jnrE;
 182             j_coord_offsetF  = DIM*jnrF;
 183             j_coord_offsetG  = DIM*jnrG;
 184             j_coord_offsetH  = DIM*jnrH;
 185
 186             /* load j atom coordinates */
 187             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 188                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 189                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 190                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 191                                                  &jx0,&jy0,&jz0);
 192
 193             /* Calculate displacement vector */
 194             dx00             = _mm256_sub_ps(ix0,jx0);
 195             dy00             = _mm256_sub_ps(iy0,jy0);
 196             dz00             = _mm256_sub_ps(iz0,jz0);
 197
 198             /* Calculate squared distance and things based on it */
 199             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 200
 201             rinvsq00         = gmx_mm256_inv_ps(rsq00);
 202
 203             /* Load parameters for j particles */
 204             vdwjidx0A        = 2*vdwtype[jnrA+0];
 205             vdwjidx0B        = 2*vdwtype[jnrB+0];
 206             vdwjidx0C        = 2*vdwtype[jnrC+0];
 207             vdwjidx0D        = 2*vdwtype[jnrD+0];
 208             vdwjidx0E        = 2*vdwtype[jnrE+0];
 209             vdwjidx0F        = 2*vdwtype[jnrF+0];
 210             vdwjidx0G        = 2*vdwtype[jnrG+0];
 211             vdwjidx0H        = 2*vdwtype[jnrH+0];
 212
 213             /**************************
 214              * CALCULATE INTERACTIONS *
 215              **************************/
 216
 217             /* Compute parameters for interactions between i and j atoms */
 218             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 219                                             vdwioffsetptr0+vdwjidx0B,
 220                                             vdwioffsetptr0+vdwjidx0C,
 221                                             vdwioffsetptr0+vdwjidx0D,
 222                                             vdwioffsetptr0+vdwjidx0E,
 223                                             vdwioffsetptr0+vdwjidx0F,
 224                                             vdwioffsetptr0+vdwjidx0G,
 225                                             vdwioffsetptr0+vdwjidx0H,
 226                                             &c6_00,&c12_00);
 227
 228             /* LENNARD-JONES DISPERSION/REPULSION */
 229
 230             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 231             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 232             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 233             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
 234             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 235
 236             /* Update potential sum for this i atom from the interaction with this j atom. */
 237             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 238
 239             fscal            = fvdw;
 240
 241             /* Calculate temporary vectorial force */
 242             tx               = _mm256_mul_ps(fscal,dx00);
 243             ty               = _mm256_mul_ps(fscal,dy00);
 244             tz               = _mm256_mul_ps(fscal,dz00);
 245
 246             /* Update vectorial force */
 247             fix0             = _mm256_add_ps(fix0,tx);
 248             fiy0             = _mm256_add_ps(fiy0,ty);
 249             fiz0             = _mm256_add_ps(fiz0,tz);
 250
 251             fjptrA             = f+j_coord_offsetA;
 252             fjptrB             = f+j_coord_offsetB;
 253             fjptrC             = f+j_coord_offsetC;
 254             fjptrD             = f+j_coord_offsetD;
 255             fjptrE             = f+j_coord_offsetE;
 256             fjptrF             = f+j_coord_offsetF;
 257             fjptrG             = f+j_coord_offsetG;
 258             fjptrH             = f+j_coord_offsetH;
 259             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 260
 261             /* Inner loop uses 32 flops */
 262         }
 263
 264         if(jidx<j_index_end)
 265         {
 266
 267             /* Get j neighbor index, and coordinate index */
 268             jnrlistA         = jjnr[jidx];
 269             jnrlistB         = jjnr[jidx+1];
 270             jnrlistC         = jjnr[jidx+2];
 271             jnrlistD         = jjnr[jidx+3];
 272             jnrlistE         = jjnr[jidx+4];
 273             jnrlistF         = jjnr[jidx+5];
 274             jnrlistG         = jjnr[jidx+6];
 275             jnrlistH         = jjnr[jidx+7];
 276             /* Sign of each element will be negative for non-real atoms.
 277              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 278              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 279              */
 280             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 281                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 282
 283             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 284             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 285             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 286             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 287             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 288             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 289             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 290             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 291             j_coord_offsetA  = DIM*jnrA;
 292             j_coord_offsetB  = DIM*jnrB;
 293             j_coord_offsetC  = DIM*jnrC;
 294             j_coord_offsetD  = DIM*jnrD;
 295             j_coord_offsetE  = DIM*jnrE;
 296             j_coord_offsetF  = DIM*jnrF;
 297             j_coord_offsetG  = DIM*jnrG;
 298             j_coord_offsetH  = DIM*jnrH;
 299
 300             /* load j atom coordinates */
 301             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 302                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 303                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 304                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 305                                                  &jx0,&jy0,&jz0);
 306
 307             /* Calculate displacement vector */
 308             dx00             = _mm256_sub_ps(ix0,jx0);
 309             dy00             = _mm256_sub_ps(iy0,jy0);
 310             dz00             = _mm256_sub_ps(iz0,jz0);
 311
 312             /* Calculate squared distance and things based on it */
 313             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 314
 315             rinvsq00         = gmx_mm256_inv_ps(rsq00);
 316
 317             /* Load parameters for j particles */
 318             vdwjidx0A        = 2*vdwtype[jnrA+0];
 319             vdwjidx0B        = 2*vdwtype[jnrB+0];
 320             vdwjidx0C        = 2*vdwtype[jnrC+0];
 321             vdwjidx0D        = 2*vdwtype[jnrD+0];
 322             vdwjidx0E        = 2*vdwtype[jnrE+0];
 323             vdwjidx0F        = 2*vdwtype[jnrF+0];
 324             vdwjidx0G        = 2*vdwtype[jnrG+0];
 325             vdwjidx0H        = 2*vdwtype[jnrH+0];
 326
 327             /**************************
 328              * CALCULATE INTERACTIONS *
 329              **************************/
 330
 331             /* Compute parameters for interactions between i and j atoms */
 332             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 333                                             vdwioffsetptr0+vdwjidx0B,
 334                                             vdwioffsetptr0+vdwjidx0C,
 335                                             vdwioffsetptr0+vdwjidx0D,
 336                                             vdwioffsetptr0+vdwjidx0E,
 337                                             vdwioffsetptr0+vdwjidx0F,
 338                                             vdwioffsetptr0+vdwjidx0G,
 339                                             vdwioffsetptr0+vdwjidx0H,
 340                                             &c6_00,&c12_00);
 341
 342             /* LENNARD-JONES DISPERSION/REPULSION */
 343
 344             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 345             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 346             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 347             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
 348             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 349
 350             /* Update potential sum for this i atom from the interaction with this j atom. */
 351             vvdw             = _mm256_andnot_ps(dummy_mask,vvdw);
 352             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 353
 354             fscal            = fvdw;
 355
 356             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 357
 358             /* Calculate temporary vectorial force */
 359             tx               = _mm256_mul_ps(fscal,dx00);
 360             ty               = _mm256_mul_ps(fscal,dy00);
 361             tz               = _mm256_mul_ps(fscal,dz00);
 362
 363             /* Update vectorial force */
 364             fix0             = _mm256_add_ps(fix0,tx);
 365             fiy0             = _mm256_add_ps(fiy0,ty);
 366             fiz0             = _mm256_add_ps(fiz0,tz);
 367
 368             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 369             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 370             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 371             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 372             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
 373             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
 374             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
 375             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
 376             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 377
 378             /* Inner loop uses 32 flops */
 379         }
 380
 381         /* End of innermost loop */
 382
 383         gmx_mm256_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 384                                                  f+i_coord_offset,fshift+i_shift_offset);
 385
 386         ggid                        = gid[iidx];
 387         /* Update potential energies */
 388         gmx_mm256_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 389
 390         /* Increment number of inner iterations */
 391         inneriter                  += j_index_end - j_index_start;
 392
 393         /* Outer loop uses 7 flops */
 394     }
 395
 396     /* Increment number of outer iterations */
 397     outeriter        += nri;
 398
 399     /* Update outer/inner flops */
 400
 401     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_VF,outeriter*7 + inneriter*32);
 402 }
 403 /*
 404  * Gromacs nonbonded kernel:   nb_kernel_ElecNone_VdwLJ_GeomP1P1_F_avx_256_single
 405  * Electrostatics interaction: None
 406  * VdW interaction:            LennardJones
 407  * Geometry:                   Particle-Particle
 408  * Calculate force/pot:        Force
 409  */
 410 void
 411 nb_kernel_ElecNone_VdwLJ_GeomP1P1_F_avx_256_single
 412                     (t_nblist                    * gmx_restrict       nlist,
 413                      rvec                        * gmx_restrict          xx,
 414                      rvec                        * gmx_restrict          ff,
 415                      t_forcerec                  * gmx_restrict          fr,
 416                      t_mdatoms                   * gmx_restrict     mdatoms,
 417                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 418                      t_nrnb                      * gmx_restrict        nrnb)
 419 {
 420     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 421      * just 0 for non-waters.
 422      * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
 423      * jnr indices corresponding to data put in the four positions in the SIMD register.
 424      */
 425     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 426     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 427     int              jnrA,jnrB,jnrC,jnrD;
 428     int              jnrE,jnrF,jnrG,jnrH;
 429     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 430     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 431     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 432     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
 433     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 434     real             rcutoff_scalar;
 435     real             *shiftvec,*fshift,*x,*f;
 436     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
 437     real             scratch[4*DIM];
 438     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 439     real *           vdwioffsetptr0;
 440     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 441     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
 442     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 443     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 444     int              nvdwtype;
 445     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 446     int              *vdwtype;
 447     real             *vdwparam;
 448     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 449     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 450     __m256           dummy_mask,cutoff_mask;
 451     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 452     __m256           one     = _mm256_set1_ps(1.0);
 453     __m256           two     = _mm256_set1_ps(2.0);
 454     x                = xx[0];
 455     f                = ff[0];
 456
 457     nri              = nlist->nri;
 458     iinr             = nlist->iinr;
 459     jindex           = nlist->jindex;
 460     jjnr             = nlist->jjnr;
 461     shiftidx         = nlist->shift;
 462     gid              = nlist->gid;
 463     shiftvec         = fr->shift_vec[0];
 464     fshift           = fr->fshift[0];
 465     nvdwtype         = fr->ntype;
 466     vdwparam         = fr->nbfp;
 467     vdwtype          = mdatoms->typeA;
 468
 469     /* Avoid stupid compiler warnings */
 470     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 471     j_coord_offsetA = 0;
 472     j_coord_offsetB = 0;
 473     j_coord_offsetC = 0;
 474     j_coord_offsetD = 0;
 475     j_coord_offsetE = 0;
 476     j_coord_offsetF = 0;
 477     j_coord_offsetG = 0;
 478     j_coord_offsetH = 0;
 479
 480     outeriter        = 0;
 481     inneriter        = 0;
 482
 483     for(iidx=0;iidx<4*DIM;iidx++)
 484     {
 485         scratch[iidx] = 0.0;
 486     }
 487
 488     /* Start outer loop over neighborlists */
 489     for(iidx=0; iidx<nri; iidx++)
 490     {
 491         /* Load shift vector for this list */
 492         i_shift_offset   = DIM*shiftidx[iidx];
 493
 494         /* Load limits for loop over neighbors */
 495         j_index_start    = jindex[iidx];
 496         j_index_end      = jindex[iidx+1];
 497
 498         /* Get outer coordinate index */
 499         inr              = iinr[iidx];
 500         i_coord_offset   = DIM*inr;
 501
 502         /* Load i particle coords and add shift vector */
 503         gmx_mm256_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 504
 505         fix0             = _mm256_setzero_ps();
 506         fiy0             = _mm256_setzero_ps();
 507         fiz0             = _mm256_setzero_ps();
 508
 509         /* Load parameters for i particles */
 510         vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 511
 512         /* Start inner kernel loop */
 513         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 514         {
 515
 516             /* Get j neighbor index, and coordinate index */
 517             jnrA             = jjnr[jidx];
 518             jnrB             = jjnr[jidx+1];
 519             jnrC             = jjnr[jidx+2];
 520             jnrD             = jjnr[jidx+3];
 521             jnrE             = jjnr[jidx+4];
 522             jnrF             = jjnr[jidx+5];
 523             jnrG             = jjnr[jidx+6];
 524             jnrH             = jjnr[jidx+7];
 525             j_coord_offsetA  = DIM*jnrA;
 526             j_coord_offsetB  = DIM*jnrB;
 527             j_coord_offsetC  = DIM*jnrC;
 528             j_coord_offsetD  = DIM*jnrD;
 529             j_coord_offsetE  = DIM*jnrE;
 530             j_coord_offsetF  = DIM*jnrF;
 531             j_coord_offsetG  = DIM*jnrG;
 532             j_coord_offsetH  = DIM*jnrH;
 533
 534             /* load j atom coordinates */
 535             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 536                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 537                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 538                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 539                                                  &jx0,&jy0,&jz0);
 540
 541             /* Calculate displacement vector */
 542             dx00             = _mm256_sub_ps(ix0,jx0);
 543             dy00             = _mm256_sub_ps(iy0,jy0);
 544             dz00             = _mm256_sub_ps(iz0,jz0);
 545
 546             /* Calculate squared distance and things based on it */
 547             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 548
 549             rinvsq00         = gmx_mm256_inv_ps(rsq00);
 550
 551             /* Load parameters for j particles */
 552             vdwjidx0A        = 2*vdwtype[jnrA+0];
 553             vdwjidx0B        = 2*vdwtype[jnrB+0];
 554             vdwjidx0C        = 2*vdwtype[jnrC+0];
 555             vdwjidx0D        = 2*vdwtype[jnrD+0];
 556             vdwjidx0E        = 2*vdwtype[jnrE+0];
 557             vdwjidx0F        = 2*vdwtype[jnrF+0];
 558             vdwjidx0G        = 2*vdwtype[jnrG+0];
 559             vdwjidx0H        = 2*vdwtype[jnrH+0];
 560
 561             /**************************
 562              * CALCULATE INTERACTIONS *
 563              **************************/
 564
 565             /* Compute parameters for interactions between i and j atoms */
 566             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 567                                             vdwioffsetptr0+vdwjidx0B,
 568                                             vdwioffsetptr0+vdwjidx0C,
 569                                             vdwioffsetptr0+vdwjidx0D,
 570                                             vdwioffsetptr0+vdwjidx0E,
 571                                             vdwioffsetptr0+vdwjidx0F,
 572                                             vdwioffsetptr0+vdwjidx0G,
 573                                             vdwioffsetptr0+vdwjidx0H,
 574                                             &c6_00,&c12_00);
 575
 576             /* LENNARD-JONES DISPERSION/REPULSION */
 577
 578             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 579             fvdw             = _mm256_mul_ps(_mm256_sub_ps(_mm256_mul_ps(c12_00,rinvsix),c6_00),_mm256_mul_ps(rinvsix,rinvsq00));
 580
 581             fscal            = fvdw;
 582
 583             /* Calculate temporary vectorial force */
 584             tx               = _mm256_mul_ps(fscal,dx00);
 585             ty               = _mm256_mul_ps(fscal,dy00);
 586             tz               = _mm256_mul_ps(fscal,dz00);
 587
 588             /* Update vectorial force */
 589             fix0             = _mm256_add_ps(fix0,tx);
 590             fiy0             = _mm256_add_ps(fiy0,ty);
 591             fiz0             = _mm256_add_ps(fiz0,tz);
 592
 593             fjptrA             = f+j_coord_offsetA;
 594             fjptrB             = f+j_coord_offsetB;
 595             fjptrC             = f+j_coord_offsetC;
 596             fjptrD             = f+j_coord_offsetD;
 597             fjptrE             = f+j_coord_offsetE;
 598             fjptrF             = f+j_coord_offsetF;
 599             fjptrG             = f+j_coord_offsetG;
 600             fjptrH             = f+j_coord_offsetH;
 601             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 602
 603             /* Inner loop uses 27 flops */
 604         }
 605
 606         if(jidx<j_index_end)
 607         {
 608
 609             /* Get j neighbor index, and coordinate index */
 610             jnrlistA         = jjnr[jidx];
 611             jnrlistB         = jjnr[jidx+1];
 612             jnrlistC         = jjnr[jidx+2];
 613             jnrlistD         = jjnr[jidx+3];
 614             jnrlistE         = jjnr[jidx+4];
 615             jnrlistF         = jjnr[jidx+5];
 616             jnrlistG         = jjnr[jidx+6];
 617             jnrlistH         = jjnr[jidx+7];
 618             /* Sign of each element will be negative for non-real atoms.
 619              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 620              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 621              */
 622             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 623                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 624
 625             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 626             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 627             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 628             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 629             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 630             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 631             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 632             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 633             j_coord_offsetA  = DIM*jnrA;
 634             j_coord_offsetB  = DIM*jnrB;
 635             j_coord_offsetC  = DIM*jnrC;
 636             j_coord_offsetD  = DIM*jnrD;
 637             j_coord_offsetE  = DIM*jnrE;
 638             j_coord_offsetF  = DIM*jnrF;
 639             j_coord_offsetG  = DIM*jnrG;
 640             j_coord_offsetH  = DIM*jnrH;
 641
 642             /* load j atom coordinates */
 643             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 644                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 645                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 646                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 647                                                  &jx0,&jy0,&jz0);
 648
 649             /* Calculate displacement vector */
 650             dx00             = _mm256_sub_ps(ix0,jx0);
 651             dy00             = _mm256_sub_ps(iy0,jy0);
 652             dz00             = _mm256_sub_ps(iz0,jz0);
 653
 654             /* Calculate squared distance and things based on it */
 655             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 656
 657             rinvsq00         = gmx_mm256_inv_ps(rsq00);
 658
 659             /* Load parameters for j particles */
 660             vdwjidx0A        = 2*vdwtype[jnrA+0];
 661             vdwjidx0B        = 2*vdwtype[jnrB+0];
 662             vdwjidx0C        = 2*vdwtype[jnrC+0];
 663             vdwjidx0D        = 2*vdwtype[jnrD+0];
 664             vdwjidx0E        = 2*vdwtype[jnrE+0];
 665             vdwjidx0F        = 2*vdwtype[jnrF+0];
 666             vdwjidx0G        = 2*vdwtype[jnrG+0];
 667             vdwjidx0H        = 2*vdwtype[jnrH+0];
 668
 669             /**************************
 670              * CALCULATE INTERACTIONS *
 671              **************************/
 672
 673             /* Compute parameters for interactions between i and j atoms */
 674             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 675                                             vdwioffsetptr0+vdwjidx0B,
 676                                             vdwioffsetptr0+vdwjidx0C,
 677                                             vdwioffsetptr0+vdwjidx0D,
 678                                             vdwioffsetptr0+vdwjidx0E,
 679                                             vdwioffsetptr0+vdwjidx0F,
 680                                             vdwioffsetptr0+vdwjidx0G,
 681                                             vdwioffsetptr0+vdwjidx0H,
 682                                             &c6_00,&c12_00);
 683
 684             /* LENNARD-JONES DISPERSION/REPULSION */
 685
 686             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 687             fvdw             = _mm256_mul_ps(_mm256_sub_ps(_mm256_mul_ps(c12_00,rinvsix),c6_00),_mm256_mul_ps(rinvsix,rinvsq00));
 688
 689             fscal            = fvdw;
 690
 691             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 692
 693             /* Calculate temporary vectorial force */
 694             tx               = _mm256_mul_ps(fscal,dx00);
 695             ty               = _mm256_mul_ps(fscal,dy00);
 696             tz               = _mm256_mul_ps(fscal,dz00);
 697
 698             /* Update vectorial force */
 699             fix0             = _mm256_add_ps(fix0,tx);
 700             fiy0             = _mm256_add_ps(fiy0,ty);
 701             fiz0             = _mm256_add_ps(fiz0,tz);
 702
 703             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 704             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 705             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 706             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 707             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
 708             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
 709             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
 710             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
 711             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 712
 713             /* Inner loop uses 27 flops */
 714         }
 715
 716         /* End of innermost loop */
 717
 718         gmx_mm256_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 719                                                  f+i_coord_offset,fshift+i_shift_offset);
 720
 721         /* Increment number of inner iterations */
 722         inneriter                  += j_index_end - j_index_start;
 723
 724         /* Outer loop uses 6 flops */
 725     }
 726
 727     /* Increment number of outer iterations */
 728     outeriter        += nri;
 729
 730     /* Update outer/inner flops */
 731
 732     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_F,outeriter*6 + inneriter*27);
 733 }