src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_single/nb_kernel_ElecCoul_VdwCSTab_GeomP1P1_avx_128_fma_single.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS avx_128_fma_single kernel generator.
  37  */
  38 #ifdef HAVE_CONFIG_H
  39 #include <config.h>
  40 #endif
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "types/simple.h"
  46 #include "vec.h"
  47 #include "nrnb.h"
  48
  49 #include "gromacs/simd/math_x86_avx_128_fma_single.h"
  50 #include "kernelutil_x86_avx_128_fma_single.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwCSTab_GeomP1P1_VF_avx_128_fma_single
  54  * Electrostatics interaction: Coulomb
  55  * VdW interaction:            CubicSplineTable
  56  * Geometry:                   Particle-Particle
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecCoul_VdwCSTab_GeomP1P1_VF_avx_128_fma_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 refer to j loop unrolling done with AVX_128, e.g. for the four 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              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  78     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  79     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  80     real             rcutoff_scalar;
  81     real             *shiftvec,*fshift,*x,*f;
  82     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  83     real             scratch[4*DIM];
  84     __m128           fscal,rcutoff,rcutoff2,jidxall;
  85     int              vdwioffset0;
  86     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  87     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  88     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  89     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  90     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
  91     real             *charge;
  92     int              nvdwtype;
  93     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  94     int              *vdwtype;
  95     real             *vdwparam;
  96     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
  97     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
  98     __m128i          vfitab;
  99     __m128i          ifour       = _mm_set1_epi32(4);
 100     __m128           rt,vfeps,twovfeps,vftabscale,Y,F,G,H,Fp,VV,FF;
 101     real             *vftab;
 102     __m128           dummy_mask,cutoff_mask;
 103     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 104     __m128           one     = _mm_set1_ps(1.0);
 105     __m128           two     = _mm_set1_ps(2.0);
 106     x                = xx[0];
 107     f                = ff[0];
 108
 109     nri              = nlist->nri;
 110     iinr             = nlist->iinr;
 111     jindex           = nlist->jindex;
 112     jjnr             = nlist->jjnr;
 113     shiftidx         = nlist->shift;
 114     gid              = nlist->gid;
 115     shiftvec         = fr->shift_vec[0];
 116     fshift           = fr->fshift[0];
 117     facel            = _mm_set1_ps(fr->epsfac);
 118     charge           = mdatoms->chargeA;
 119     nvdwtype         = fr->ntype;
 120     vdwparam         = fr->nbfp;
 121     vdwtype          = mdatoms->typeA;
 122
 123     vftab            = kernel_data->table_vdw->data;
 124     vftabscale       = _mm_set1_ps(kernel_data->table_vdw->scale);
 125
 126     /* Avoid stupid compiler warnings */
 127     jnrA = jnrB = jnrC = jnrD = 0;
 128     j_coord_offsetA = 0;
 129     j_coord_offsetB = 0;
 130     j_coord_offsetC = 0;
 131     j_coord_offsetD = 0;
 132
 133     outeriter        = 0;
 134     inneriter        = 0;
 135
 136     for(iidx=0;iidx<4*DIM;iidx++)
 137     {
 138         scratch[iidx] = 0.0;
 139     }
 140
 141     /* Start outer loop over neighborlists */
 142     for(iidx=0; iidx<nri; iidx++)
 143     {
 144         /* Load shift vector for this list */
 145         i_shift_offset   = DIM*shiftidx[iidx];
 146
 147         /* Load limits for loop over neighbors */
 148         j_index_start    = jindex[iidx];
 149         j_index_end      = jindex[iidx+1];
 150
 151         /* Get outer coordinate index */
 152         inr              = iinr[iidx];
 153         i_coord_offset   = DIM*inr;
 154
 155         /* Load i particle coords and add shift vector */
 156         gmx_mm_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 157
 158         fix0             = _mm_setzero_ps();
 159         fiy0             = _mm_setzero_ps();
 160         fiz0             = _mm_setzero_ps();
 161
 162         /* Load parameters for i particles */
 163         iq0              = _mm_mul_ps(facel,_mm_load1_ps(charge+inr+0));
 164         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 165
 166         /* Reset potential sums */
 167         velecsum         = _mm_setzero_ps();
 168         vvdwsum          = _mm_setzero_ps();
 169
 170         /* Start inner kernel loop */
 171         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 172         {
 173
 174             /* Get j neighbor index, and coordinate index */
 175             jnrA             = jjnr[jidx];
 176             jnrB             = jjnr[jidx+1];
 177             jnrC             = jjnr[jidx+2];
 178             jnrD             = jjnr[jidx+3];
 179             j_coord_offsetA  = DIM*jnrA;
 180             j_coord_offsetB  = DIM*jnrB;
 181             j_coord_offsetC  = DIM*jnrC;
 182             j_coord_offsetD  = DIM*jnrD;
 183
 184             /* load j atom coordinates */
 185             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 186                                               x+j_coord_offsetC,x+j_coord_offsetD,
 187                                               &jx0,&jy0,&jz0);
 188
 189             /* Calculate displacement vector */
 190             dx00             = _mm_sub_ps(ix0,jx0);
 191             dy00             = _mm_sub_ps(iy0,jy0);
 192             dz00             = _mm_sub_ps(iz0,jz0);
 193
 194             /* Calculate squared distance and things based on it */
 195             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 196
 197             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 198
 199             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 200
 201             /* Load parameters for j particles */
 202             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 203                                                               charge+jnrC+0,charge+jnrD+0);
 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
 209             /**************************
 210              * CALCULATE INTERACTIONS *
 211              **************************/
 212
 213             r00              = _mm_mul_ps(rsq00,rinv00);
 214
 215             /* Compute parameters for interactions between i and j atoms */
 216             qq00             = _mm_mul_ps(iq0,jq0);
 217             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 218                                          vdwparam+vdwioffset0+vdwjidx0B,
 219                                          vdwparam+vdwioffset0+vdwjidx0C,
 220                                          vdwparam+vdwioffset0+vdwjidx0D,
 221                                          &c6_00,&c12_00);
 222
 223             /* Calculate table index by multiplying r with table scale and truncate to integer */
 224             rt               = _mm_mul_ps(r00,vftabscale);
 225             vfitab           = _mm_cvttps_epi32(rt);
 226 #ifdef __XOP__
 227             vfeps            = _mm_frcz_ps(rt);
 228 #else
 229             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 230 #endif
 231             twovfeps         = _mm_add_ps(vfeps,vfeps);
 232             vfitab           = _mm_slli_epi32(vfitab,3);
 233
 234             /* COULOMB ELECTROSTATICS */
 235             velec            = _mm_mul_ps(qq00,rinv00);
 236             felec            = _mm_mul_ps(velec,rinvsq00);
 237
 238             /* CUBIC SPLINE TABLE DISPERSION */
 239             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 240             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 241             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 242             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 243             _MM_TRANSPOSE4_PS(Y,F,G,H);
 244             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 245             VV               = _mm_macc_ps(vfeps,Fp,Y);
 246             vvdw6            = _mm_mul_ps(c6_00,VV);
 247             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 248             fvdw6            = _mm_mul_ps(c6_00,FF);
 249
 250             /* CUBIC SPLINE TABLE REPULSION */
 251             vfitab           = _mm_add_epi32(vfitab,ifour);
 252             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 253             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 254             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 255             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 256             _MM_TRANSPOSE4_PS(Y,F,G,H);
 257             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 258             VV               = _mm_macc_ps(vfeps,Fp,Y);
 259             vvdw12           = _mm_mul_ps(c12_00,VV);
 260             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 261             fvdw12           = _mm_mul_ps(c12_00,FF);
 262             vvdw             = _mm_add_ps(vvdw12,vvdw6);
 263             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 264
 265             /* Update potential sum for this i atom from the interaction with this j atom. */
 266             velecsum         = _mm_add_ps(velecsum,velec);
 267             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 268
 269             fscal            = _mm_add_ps(felec,fvdw);
 270
 271              /* Update vectorial force */
 272             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 273             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 274             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 275
 276             fjptrA             = f+j_coord_offsetA;
 277             fjptrB             = f+j_coord_offsetB;
 278             fjptrC             = f+j_coord_offsetC;
 279             fjptrD             = f+j_coord_offsetD;
 280             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 281                                                    _mm_mul_ps(dx00,fscal),
 282                                                    _mm_mul_ps(dy00,fscal),
 283                                                    _mm_mul_ps(dz00,fscal));
 284
 285             /* Inner loop uses 66 flops */
 286         }
 287
 288         if(jidx<j_index_end)
 289         {
 290
 291             /* Get j neighbor index, and coordinate index */
 292             jnrlistA         = jjnr[jidx];
 293             jnrlistB         = jjnr[jidx+1];
 294             jnrlistC         = jjnr[jidx+2];
 295             jnrlistD         = jjnr[jidx+3];
 296             /* Sign of each element will be negative for non-real atoms.
 297              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 298              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 299              */
 300             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 301             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 302             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 303             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 304             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 305             j_coord_offsetA  = DIM*jnrA;
 306             j_coord_offsetB  = DIM*jnrB;
 307             j_coord_offsetC  = DIM*jnrC;
 308             j_coord_offsetD  = DIM*jnrD;
 309
 310             /* load j atom coordinates */
 311             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 312                                               x+j_coord_offsetC,x+j_coord_offsetD,
 313                                               &jx0,&jy0,&jz0);
 314
 315             /* Calculate displacement vector */
 316             dx00             = _mm_sub_ps(ix0,jx0);
 317             dy00             = _mm_sub_ps(iy0,jy0);
 318             dz00             = _mm_sub_ps(iz0,jz0);
 319
 320             /* Calculate squared distance and things based on it */
 321             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 322
 323             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 324
 325             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 326
 327             /* Load parameters for j particles */
 328             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 329                                                               charge+jnrC+0,charge+jnrD+0);
 330             vdwjidx0A        = 2*vdwtype[jnrA+0];
 331             vdwjidx0B        = 2*vdwtype[jnrB+0];
 332             vdwjidx0C        = 2*vdwtype[jnrC+0];
 333             vdwjidx0D        = 2*vdwtype[jnrD+0];
 334
 335             /**************************
 336              * CALCULATE INTERACTIONS *
 337              **************************/
 338
 339             r00              = _mm_mul_ps(rsq00,rinv00);
 340             r00              = _mm_andnot_ps(dummy_mask,r00);
 341
 342             /* Compute parameters for interactions between i and j atoms */
 343             qq00             = _mm_mul_ps(iq0,jq0);
 344             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 345                                          vdwparam+vdwioffset0+vdwjidx0B,
 346                                          vdwparam+vdwioffset0+vdwjidx0C,
 347                                          vdwparam+vdwioffset0+vdwjidx0D,
 348                                          &c6_00,&c12_00);
 349
 350             /* Calculate table index by multiplying r with table scale and truncate to integer */
 351             rt               = _mm_mul_ps(r00,vftabscale);
 352             vfitab           = _mm_cvttps_epi32(rt);
 353 #ifdef __XOP__
 354             vfeps            = _mm_frcz_ps(rt);
 355 #else
 356             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 357 #endif
 358             twovfeps         = _mm_add_ps(vfeps,vfeps);
 359             vfitab           = _mm_slli_epi32(vfitab,3);
 360
 361             /* COULOMB ELECTROSTATICS */
 362             velec            = _mm_mul_ps(qq00,rinv00);
 363             felec            = _mm_mul_ps(velec,rinvsq00);
 364
 365             /* CUBIC SPLINE TABLE DISPERSION */
 366             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 367             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 368             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 369             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 370             _MM_TRANSPOSE4_PS(Y,F,G,H);
 371             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 372             VV               = _mm_macc_ps(vfeps,Fp,Y);
 373             vvdw6            = _mm_mul_ps(c6_00,VV);
 374             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 375             fvdw6            = _mm_mul_ps(c6_00,FF);
 376
 377             /* CUBIC SPLINE TABLE REPULSION */
 378             vfitab           = _mm_add_epi32(vfitab,ifour);
 379             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 380             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 381             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 382             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 383             _MM_TRANSPOSE4_PS(Y,F,G,H);
 384             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 385             VV               = _mm_macc_ps(vfeps,Fp,Y);
 386             vvdw12           = _mm_mul_ps(c12_00,VV);
 387             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 388             fvdw12           = _mm_mul_ps(c12_00,FF);
 389             vvdw             = _mm_add_ps(vvdw12,vvdw6);
 390             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 391
 392             /* Update potential sum for this i atom from the interaction with this j atom. */
 393             velec            = _mm_andnot_ps(dummy_mask,velec);
 394             velecsum         = _mm_add_ps(velecsum,velec);
 395             vvdw             = _mm_andnot_ps(dummy_mask,vvdw);
 396             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 397
 398             fscal            = _mm_add_ps(felec,fvdw);
 399
 400             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 401
 402              /* Update vectorial force */
 403             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 404             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 405             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 406
 407             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 408             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 409             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 410             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 411             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 412                                                    _mm_mul_ps(dx00,fscal),
 413                                                    _mm_mul_ps(dy00,fscal),
 414                                                    _mm_mul_ps(dz00,fscal));
 415
 416             /* Inner loop uses 67 flops */
 417         }
 418
 419         /* End of innermost loop */
 420
 421         gmx_mm_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 422                                               f+i_coord_offset,fshift+i_shift_offset);
 423
 424         ggid                        = gid[iidx];
 425         /* Update potential energies */
 426         gmx_mm_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 427         gmx_mm_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 428
 429         /* Increment number of inner iterations */
 430         inneriter                  += j_index_end - j_index_start;
 431
 432         /* Outer loop uses 9 flops */
 433     }
 434
 435     /* Increment number of outer iterations */
 436     outeriter        += nri;
 437
 438     /* Update outer/inner flops */
 439
 440     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*9 + inneriter*67);
 441 }
 442 /*
 443  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwCSTab_GeomP1P1_F_avx_128_fma_single
 444  * Electrostatics interaction: Coulomb
 445  * VdW interaction:            CubicSplineTable
 446  * Geometry:                   Particle-Particle
 447  * Calculate force/pot:        Force
 448  */
 449 void
 450 nb_kernel_ElecCoul_VdwCSTab_GeomP1P1_F_avx_128_fma_single
 451                     (t_nblist                    * gmx_restrict       nlist,
 452                      rvec                        * gmx_restrict          xx,
 453                      rvec                        * gmx_restrict          ff,
 454                      t_forcerec                  * gmx_restrict          fr,
 455                      t_mdatoms                   * gmx_restrict     mdatoms,
 456                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 457                      t_nrnb                      * gmx_restrict        nrnb)
 458 {
 459     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 460      * just 0 for non-waters.
 461      * Suffixes A,B,C,D refer to j loop unrolling done with AVX_128, e.g. for the four different
 462      * jnr indices corresponding to data put in the four positions in the SIMD register.
 463      */
 464     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 465     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 466     int              jnrA,jnrB,jnrC,jnrD;
 467     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 468     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 469     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 470     real             rcutoff_scalar;
 471     real             *shiftvec,*fshift,*x,*f;
 472     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 473     real             scratch[4*DIM];
 474     __m128           fscal,rcutoff,rcutoff2,jidxall;
 475     int              vdwioffset0;
 476     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 477     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 478     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 479     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 480     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 481     real             *charge;
 482     int              nvdwtype;
 483     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 484     int              *vdwtype;
 485     real             *vdwparam;
 486     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 487     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 488     __m128i          vfitab;
 489     __m128i          ifour       = _mm_set1_epi32(4);
 490     __m128           rt,vfeps,twovfeps,vftabscale,Y,F,G,H,Fp,VV,FF;
 491     real             *vftab;
 492     __m128           dummy_mask,cutoff_mask;
 493     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 494     __m128           one     = _mm_set1_ps(1.0);
 495     __m128           two     = _mm_set1_ps(2.0);
 496     x                = xx[0];
 497     f                = ff[0];
 498
 499     nri              = nlist->nri;
 500     iinr             = nlist->iinr;
 501     jindex           = nlist->jindex;
 502     jjnr             = nlist->jjnr;
 503     shiftidx         = nlist->shift;
 504     gid              = nlist->gid;
 505     shiftvec         = fr->shift_vec[0];
 506     fshift           = fr->fshift[0];
 507     facel            = _mm_set1_ps(fr->epsfac);
 508     charge           = mdatoms->chargeA;
 509     nvdwtype         = fr->ntype;
 510     vdwparam         = fr->nbfp;
 511     vdwtype          = mdatoms->typeA;
 512
 513     vftab            = kernel_data->table_vdw->data;
 514     vftabscale       = _mm_set1_ps(kernel_data->table_vdw->scale);
 515
 516     /* Avoid stupid compiler warnings */
 517     jnrA = jnrB = jnrC = jnrD = 0;
 518     j_coord_offsetA = 0;
 519     j_coord_offsetB = 0;
 520     j_coord_offsetC = 0;
 521     j_coord_offsetD = 0;
 522
 523     outeriter        = 0;
 524     inneriter        = 0;
 525
 526     for(iidx=0;iidx<4*DIM;iidx++)
 527     {
 528         scratch[iidx] = 0.0;
 529     }
 530
 531     /* Start outer loop over neighborlists */
 532     for(iidx=0; iidx<nri; iidx++)
 533     {
 534         /* Load shift vector for this list */
 535         i_shift_offset   = DIM*shiftidx[iidx];
 536
 537         /* Load limits for loop over neighbors */
 538         j_index_start    = jindex[iidx];
 539         j_index_end      = jindex[iidx+1];
 540
 541         /* Get outer coordinate index */
 542         inr              = iinr[iidx];
 543         i_coord_offset   = DIM*inr;
 544
 545         /* Load i particle coords and add shift vector */
 546         gmx_mm_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 547
 548         fix0             = _mm_setzero_ps();
 549         fiy0             = _mm_setzero_ps();
 550         fiz0             = _mm_setzero_ps();
 551
 552         /* Load parameters for i particles */
 553         iq0              = _mm_mul_ps(facel,_mm_load1_ps(charge+inr+0));
 554         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 555
 556         /* Start inner kernel loop */
 557         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 558         {
 559
 560             /* Get j neighbor index, and coordinate index */
 561             jnrA             = jjnr[jidx];
 562             jnrB             = jjnr[jidx+1];
 563             jnrC             = jjnr[jidx+2];
 564             jnrD             = jjnr[jidx+3];
 565             j_coord_offsetA  = DIM*jnrA;
 566             j_coord_offsetB  = DIM*jnrB;
 567             j_coord_offsetC  = DIM*jnrC;
 568             j_coord_offsetD  = DIM*jnrD;
 569
 570             /* load j atom coordinates */
 571             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 572                                               x+j_coord_offsetC,x+j_coord_offsetD,
 573                                               &jx0,&jy0,&jz0);
 574
 575             /* Calculate displacement vector */
 576             dx00             = _mm_sub_ps(ix0,jx0);
 577             dy00             = _mm_sub_ps(iy0,jy0);
 578             dz00             = _mm_sub_ps(iz0,jz0);
 579
 580             /* Calculate squared distance and things based on it */
 581             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 582
 583             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 584
 585             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 586
 587             /* Load parameters for j particles */
 588             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 589                                                               charge+jnrC+0,charge+jnrD+0);
 590             vdwjidx0A        = 2*vdwtype[jnrA+0];
 591             vdwjidx0B        = 2*vdwtype[jnrB+0];
 592             vdwjidx0C        = 2*vdwtype[jnrC+0];
 593             vdwjidx0D        = 2*vdwtype[jnrD+0];
 594
 595             /**************************
 596              * CALCULATE INTERACTIONS *
 597              **************************/
 598
 599             r00              = _mm_mul_ps(rsq00,rinv00);
 600
 601             /* Compute parameters for interactions between i and j atoms */
 602             qq00             = _mm_mul_ps(iq0,jq0);
 603             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 604                                          vdwparam+vdwioffset0+vdwjidx0B,
 605                                          vdwparam+vdwioffset0+vdwjidx0C,
 606                                          vdwparam+vdwioffset0+vdwjidx0D,
 607                                          &c6_00,&c12_00);
 608
 609             /* Calculate table index by multiplying r with table scale and truncate to integer */
 610             rt               = _mm_mul_ps(r00,vftabscale);
 611             vfitab           = _mm_cvttps_epi32(rt);
 612 #ifdef __XOP__
 613             vfeps            = _mm_frcz_ps(rt);
 614 #else
 615             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 616 #endif
 617             twovfeps         = _mm_add_ps(vfeps,vfeps);
 618             vfitab           = _mm_slli_epi32(vfitab,3);
 619
 620             /* COULOMB ELECTROSTATICS */
 621             velec            = _mm_mul_ps(qq00,rinv00);
 622             felec            = _mm_mul_ps(velec,rinvsq00);
 623
 624             /* CUBIC SPLINE TABLE DISPERSION */
 625             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 626             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 627             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 628             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 629             _MM_TRANSPOSE4_PS(Y,F,G,H);
 630             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 631             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 632             fvdw6            = _mm_mul_ps(c6_00,FF);
 633
 634             /* CUBIC SPLINE TABLE REPULSION */
 635             vfitab           = _mm_add_epi32(vfitab,ifour);
 636             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 637             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 638             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 639             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 640             _MM_TRANSPOSE4_PS(Y,F,G,H);
 641             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 642             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 643             fvdw12           = _mm_mul_ps(c12_00,FF);
 644             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 645
 646             fscal            = _mm_add_ps(felec,fvdw);
 647
 648              /* Update vectorial force */
 649             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 650             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 651             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 652
 653             fjptrA             = f+j_coord_offsetA;
 654             fjptrB             = f+j_coord_offsetB;
 655             fjptrC             = f+j_coord_offsetC;
 656             fjptrD             = f+j_coord_offsetD;
 657             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 658                                                    _mm_mul_ps(dx00,fscal),
 659                                                    _mm_mul_ps(dy00,fscal),
 660                                                    _mm_mul_ps(dz00,fscal));
 661
 662             /* Inner loop uses 57 flops */
 663         }
 664
 665         if(jidx<j_index_end)
 666         {
 667
 668             /* Get j neighbor index, and coordinate index */
 669             jnrlistA         = jjnr[jidx];
 670             jnrlistB         = jjnr[jidx+1];
 671             jnrlistC         = jjnr[jidx+2];
 672             jnrlistD         = jjnr[jidx+3];
 673             /* Sign of each element will be negative for non-real atoms.
 674              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 675              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 676              */
 677             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 678             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 679             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 680             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 681             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 682             j_coord_offsetA  = DIM*jnrA;
 683             j_coord_offsetB  = DIM*jnrB;
 684             j_coord_offsetC  = DIM*jnrC;
 685             j_coord_offsetD  = DIM*jnrD;
 686
 687             /* load j atom coordinates */
 688             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 689                                               x+j_coord_offsetC,x+j_coord_offsetD,
 690                                               &jx0,&jy0,&jz0);
 691
 692             /* Calculate displacement vector */
 693             dx00             = _mm_sub_ps(ix0,jx0);
 694             dy00             = _mm_sub_ps(iy0,jy0);
 695             dz00             = _mm_sub_ps(iz0,jz0);
 696
 697             /* Calculate squared distance and things based on it */
 698             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 699
 700             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 701
 702             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 703
 704             /* Load parameters for j particles */
 705             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 706                                                               charge+jnrC+0,charge+jnrD+0);
 707             vdwjidx0A        = 2*vdwtype[jnrA+0];
 708             vdwjidx0B        = 2*vdwtype[jnrB+0];
 709             vdwjidx0C        = 2*vdwtype[jnrC+0];
 710             vdwjidx0D        = 2*vdwtype[jnrD+0];
 711
 712             /**************************
 713              * CALCULATE INTERACTIONS *
 714              **************************/
 715
 716             r00              = _mm_mul_ps(rsq00,rinv00);
 717             r00              = _mm_andnot_ps(dummy_mask,r00);
 718
 719             /* Compute parameters for interactions between i and j atoms */
 720             qq00             = _mm_mul_ps(iq0,jq0);
 721             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 722                                          vdwparam+vdwioffset0+vdwjidx0B,
 723                                          vdwparam+vdwioffset0+vdwjidx0C,
 724                                          vdwparam+vdwioffset0+vdwjidx0D,
 725                                          &c6_00,&c12_00);
 726
 727             /* Calculate table index by multiplying r with table scale and truncate to integer */
 728             rt               = _mm_mul_ps(r00,vftabscale);
 729             vfitab           = _mm_cvttps_epi32(rt);
 730 #ifdef __XOP__
 731             vfeps            = _mm_frcz_ps(rt);
 732 #else
 733             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 734 #endif
 735             twovfeps         = _mm_add_ps(vfeps,vfeps);
 736             vfitab           = _mm_slli_epi32(vfitab,3);
 737
 738             /* COULOMB ELECTROSTATICS */
 739             velec            = _mm_mul_ps(qq00,rinv00);
 740             felec            = _mm_mul_ps(velec,rinvsq00);
 741
 742             /* CUBIC SPLINE TABLE DISPERSION */
 743             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 744             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 745             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 746             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 747             _MM_TRANSPOSE4_PS(Y,F,G,H);
 748             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 749             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 750             fvdw6            = _mm_mul_ps(c6_00,FF);
 751
 752             /* CUBIC SPLINE TABLE REPULSION */
 753             vfitab           = _mm_add_epi32(vfitab,ifour);
 754             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 755             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 756             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 757             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 758             _MM_TRANSPOSE4_PS(Y,F,G,H);
 759             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 760             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 761             fvdw12           = _mm_mul_ps(c12_00,FF);
 762             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 763
 764             fscal            = _mm_add_ps(felec,fvdw);
 765
 766             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 767
 768              /* Update vectorial force */
 769             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 770             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 771             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 772
 773             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 774             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 775             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 776             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 777             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 778                                                    _mm_mul_ps(dx00,fscal),
 779                                                    _mm_mul_ps(dy00,fscal),
 780                                                    _mm_mul_ps(dz00,fscal));
 781
 782             /* Inner loop uses 58 flops */
 783         }
 784
 785         /* End of innermost loop */
 786
 787         gmx_mm_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 788                                               f+i_coord_offset,fshift+i_shift_offset);
 789
 790         /* Increment number of inner iterations */
 791         inneriter                  += j_index_end - j_index_start;
 792
 793         /* Outer loop uses 7 flops */
 794     }
 795
 796     /* Increment number of outer iterations */
 797     outeriter        += nri;
 798
 799     /* Update outer/inner flops */
 800
 801     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*7 + inneriter*58);
 802 }