src/gromacs/gmxlib/nonbonded/nb_kernel_sse2_double/kernelutil_x86_sse2_double.h

   1 /*
   2  *                This source code is part of
   3  *
   4  *                 G   R   O   M   A   C   S
   5  *
   6  * Copyright (c) 2011-2012, The GROMACS Development Team
   7  *
   8  * Gromacs is a library for molecular simulation and trajectory analysis,
   9  * written by Erik Lindahl, David van der Spoel, Berk Hess, and others - for
  10  * a full list of developers and information, check out http://www.gromacs.org
  11  *
  12  * This program is free software; you can redistribute it and/or modify it under
  13  * the terms of the GNU Lesser General Public License as published by the Free
  14  * Software Foundation; either version 2 of the License, or (at your option) any
  15  * later version.
  16  * As a special exception, you may use this file as part of a free software
  17  * library without restriction.  Specifically, if other files instantiate
  18  * templates or use macros or inline functions from this file, or you compile
  19  * this file and link it with other files to produce an executable, this
  20  * file does not by itself cause the resulting executable to be covered by
  21  * the GNU Lesser General Public License.
  22  *
  23  * In plain-speak: do not worry about classes/macros/templates either - only
  24  * changes to the library have to be LGPL, not an application linking with it.
  25  *
  26  * To help fund GROMACS development, we humbly ask that you cite
  27  * the papers people have written on it - you can find them on the website!
  28  */
  29 #ifndef _kernelutil_x86_sse2_double_h_
  30 #define _kernelutil_x86_sse2_double_h_
  31
  32 #include <math.h>
  33
  34 #include "gmx_x86_sse2.h"
  35
  36 #include <stdio.h>
  37
  38
  39 /* Normal sum of four ymm registers */
  40 #define gmx_mm_sum4_pd(t0,t1,t2,t3)  _mm_add_pd(_mm_add_pd(t0,t1),_mm_add_pd(t2,t3))
  41
  42 static int
  43 gmx_mm_any_lt(__m128d a, __m128d b)
  44 {
  45     return _mm_movemask_pd(_mm_cmplt_pd(a,b));
  46 }
  47
  48 static gmx_inline __m128d
  49 gmx_mm_calc_rsq_pd(__m128d dx, __m128d dy, __m128d dz)
  50 {
  51     return _mm_add_pd( _mm_add_pd( _mm_mul_pd(dx,dx), _mm_mul_pd(dy,dy) ), _mm_mul_pd(dz,dz) );
  52 }
  53
  54
  55 /* Load a double value from 1-2 places, merge into xmm register */
  56 static gmx_inline __m128d
  57 gmx_mm_load_2real_swizzle_pd(const double * gmx_restrict ptrA,
  58                              const double * gmx_restrict ptrB)
  59 {
  60     return _mm_unpacklo_pd(_mm_load_sd(ptrA),_mm_load_sd(ptrB));
  61 }
  62
  63 static gmx_inline __m128d
  64 gmx_mm_load_1real_pd(const double * gmx_restrict ptrA)
  65 {
  66     return _mm_load_sd(ptrA);
  67 }
  68
  69
  70 static gmx_inline void
  71 gmx_mm_store_2real_swizzle_pd(double * gmx_restrict ptrA,
  72                               double * gmx_restrict ptrB,
  73                               __m128d xmm1)
  74 {
  75     __m128d t2;
  76
  77     t2       = _mm_unpackhi_pd(xmm1,xmm1);
  78     _mm_store_sd(ptrA,xmm1);
  79     _mm_store_sd(ptrB,t2);
  80 }
  81
  82 static gmx_inline void
  83 gmx_mm_store_1real_pd(double * gmx_restrict ptrA, __m128d xmm1)
  84 {
  85     _mm_store_sd(ptrA,xmm1);
  86 }
  87
  88
  89 /* Similar to store, but increments value in memory */
  90 static gmx_inline void
  91 gmx_mm_increment_2real_swizzle_pd(double * gmx_restrict ptrA,
  92                                   double * gmx_restrict ptrB, __m128d xmm1)
  93 {
  94     __m128d t1;
  95
  96     t1   = _mm_unpackhi_pd(xmm1,xmm1);
  97     xmm1 = _mm_add_sd(xmm1,_mm_load_sd(ptrA));
  98     t1   = _mm_add_sd(t1,_mm_load_sd(ptrB));
  99     _mm_store_sd(ptrA,xmm1);
 100     _mm_store_sd(ptrB,t1);
 101 }
 102
 103 static gmx_inline void
 104 gmx_mm_increment_1real_pd(double * gmx_restrict ptrA, __m128d xmm1)
 105 {
 106     __m128d tmp;
 107
 108     tmp = gmx_mm_load_1real_pd(ptrA);
 109     tmp = _mm_add_sd(tmp,xmm1);
 110     gmx_mm_store_1real_pd(ptrA,tmp);
 111 }
 112
 113
 114 static gmx_inline void
 115 gmx_mm_load_2pair_swizzle_pd(const double * gmx_restrict p1,
 116                              const double * gmx_restrict p2,
 117                              __m128d * gmx_restrict c6,
 118                              __m128d * gmx_restrict c12)
 119 {
 120     __m128d t1,t2,t3;
 121
 122     t1   = _mm_loadu_pd(p1);
 123     t2   = _mm_loadu_pd(p2);
 124     *c6  = _mm_unpacklo_pd(t1,t2);
 125     *c12 = _mm_unpackhi_pd(t1,t2);
 126 }
 127
 128 static gmx_inline void
 129 gmx_mm_load_1pair_swizzle_pd(const double * gmx_restrict p1,
 130                              __m128d * gmx_restrict c6,
 131                              __m128d * gmx_restrict c12)
 132 {
 133     *c6     = _mm_load_sd(p1);
 134     *c12    = _mm_load_sd(p1+1);
 135 }
 136
 137
 138
 139 static gmx_inline void
 140 gmx_mm_load_shift_and_1rvec_broadcast_pd(const double * gmx_restrict xyz_shift,
 141         const double * gmx_restrict xyz,
 142         __m128d * gmx_restrict x1,
 143         __m128d * gmx_restrict y1,
 144         __m128d * gmx_restrict z1)
 145 {
 146     __m128d mem_xy,mem_z,mem_sxy,mem_sz;
 147
 148     mem_xy  = _mm_loadu_pd(xyz);
 149     mem_z   = _mm_load_sd(xyz+2);
 150     mem_sxy = _mm_loadu_pd(xyz_shift);
 151     mem_sz  = _mm_load_sd(xyz_shift+2);
 152
 153     mem_xy  = _mm_add_pd(mem_xy,mem_sxy);
 154     mem_z   = _mm_add_pd(mem_z,mem_sz);
 155
 156     *x1  = _mm_shuffle_pd(mem_xy,mem_xy,_MM_SHUFFLE2(0,0));
 157     *y1  = _mm_shuffle_pd(mem_xy,mem_xy,_MM_SHUFFLE2(1,1));
 158     *z1  = _mm_shuffle_pd(mem_z,mem_z,_MM_SHUFFLE2(0,0));
 159 }
 160
 161
 162 static gmx_inline void
 163 gmx_mm_load_shift_and_3rvec_broadcast_pd(const double * gmx_restrict xyz_shift,
 164         const double * gmx_restrict xyz,
 165         __m128d * gmx_restrict x1, __m128d * gmx_restrict y1, __m128d * gmx_restrict z1,
 166         __m128d * gmx_restrict x2, __m128d * gmx_restrict y2, __m128d * gmx_restrict z2,
 167         __m128d * gmx_restrict x3, __m128d * gmx_restrict y3, __m128d * gmx_restrict z3)
 168 {
 169     __m128d t1,t2,t3,t4,t5,sxy,sz,szx,syz;
 170
 171     t1  = _mm_loadu_pd(xyz);
 172     t2  = _mm_loadu_pd(xyz+2);
 173     t3  = _mm_loadu_pd(xyz+4);
 174     t4  = _mm_loadu_pd(xyz+6);
 175     t5  = _mm_load_sd(xyz+8);
 176
 177     sxy = _mm_loadu_pd(xyz_shift);
 178     sz  = _mm_load_sd(xyz_shift+2);
 179     szx = _mm_shuffle_pd(sz,sxy,_MM_SHUFFLE2(0,0));
 180     syz = _mm_shuffle_pd(sxy,sz,_MM_SHUFFLE2(0,1));
 181
 182     t1  = _mm_add_pd(t1,sxy);
 183     t2  = _mm_add_pd(t2,szx);
 184     t3  = _mm_add_pd(t3,syz);
 185     t4  = _mm_add_pd(t4,sxy);
 186     t5  = _mm_add_sd(t5,sz);
 187
 188     *x1  = _mm_shuffle_pd(t1,t1,_MM_SHUFFLE2(0,0));
 189     *y1  = _mm_shuffle_pd(t1,t1,_MM_SHUFFLE2(1,1));
 190     *z1  = _mm_shuffle_pd(t2,t2,_MM_SHUFFLE2(0,0));
 191     *x2  = _mm_shuffle_pd(t2,t2,_MM_SHUFFLE2(1,1));
 192     *y2  = _mm_shuffle_pd(t3,t3,_MM_SHUFFLE2(0,0));
 193     *z2  = _mm_shuffle_pd(t3,t3,_MM_SHUFFLE2(1,1));
 194     *x3  = _mm_shuffle_pd(t4,t4,_MM_SHUFFLE2(0,0));
 195     *y3  = _mm_shuffle_pd(t4,t4,_MM_SHUFFLE2(1,1));
 196     *z3  = _mm_shuffle_pd(t5,t5,_MM_SHUFFLE2(0,0));
 197 }
 198
 199
 200 static gmx_inline void
 201 gmx_mm_load_shift_and_4rvec_broadcast_pd(const double * gmx_restrict xyz_shift,
 202         const double * gmx_restrict xyz,
 203         __m128d * gmx_restrict x1, __m128d * gmx_restrict y1, __m128d * gmx_restrict z1,
 204         __m128d * gmx_restrict x2, __m128d * gmx_restrict y2, __m128d * gmx_restrict z2,
 205         __m128d * gmx_restrict x3, __m128d * gmx_restrict y3, __m128d * gmx_restrict z3,
 206         __m128d * gmx_restrict x4, __m128d * gmx_restrict y4, __m128d * gmx_restrict z4)
 207 {
 208     __m128d t1,t2,t3,t4,t5,t6,sxy,sz,szx,syz;
 209
 210     t1  = _mm_loadu_pd(xyz);
 211     t2  = _mm_loadu_pd(xyz+2);
 212     t3  = _mm_loadu_pd(xyz+4);
 213     t4  = _mm_loadu_pd(xyz+6);
 214     t5  = _mm_loadu_pd(xyz+8);
 215     t6  = _mm_loadu_pd(xyz+10);
 216
 217     sxy = _mm_loadu_pd(xyz_shift);
 218     sz  = _mm_load_sd(xyz_shift+2);
 219     szx = _mm_shuffle_pd(sz,sxy,_MM_SHUFFLE2(0,0));
 220     syz = _mm_shuffle_pd(sxy,sz,_MM_SHUFFLE2(0,1));
 221
 222     t1  = _mm_add_pd(t1,sxy);
 223     t2  = _mm_add_pd(t2,szx);
 224     t3  = _mm_add_pd(t3,syz);
 225     t4  = _mm_add_pd(t4,sxy);
 226     t5  = _mm_add_pd(t5,szx);
 227     t6  = _mm_add_pd(t6,syz);
 228
 229     *x1  = _mm_shuffle_pd(t1,t1,_MM_SHUFFLE2(0,0));
 230     *y1  = _mm_shuffle_pd(t1,t1,_MM_SHUFFLE2(1,1));
 231     *z1  = _mm_shuffle_pd(t2,t2,_MM_SHUFFLE2(0,0));
 232     *x2  = _mm_shuffle_pd(t2,t2,_MM_SHUFFLE2(1,1));
 233     *y2  = _mm_shuffle_pd(t3,t3,_MM_SHUFFLE2(0,0));
 234     *z2  = _mm_shuffle_pd(t3,t3,_MM_SHUFFLE2(1,1));
 235     *x3  = _mm_shuffle_pd(t4,t4,_MM_SHUFFLE2(0,0));
 236     *y3  = _mm_shuffle_pd(t4,t4,_MM_SHUFFLE2(1,1));
 237     *z3  = _mm_shuffle_pd(t5,t5,_MM_SHUFFLE2(0,0));
 238     *x4  = _mm_shuffle_pd(t5,t5,_MM_SHUFFLE2(1,1));
 239     *y4  = _mm_shuffle_pd(t6,t6,_MM_SHUFFLE2(0,0));
 240     *z4  = _mm_shuffle_pd(t6,t6,_MM_SHUFFLE2(1,1));
 241 }
 242
 243
 244
 245
 246 static gmx_inline void
 247 gmx_mm_load_1rvec_1ptr_swizzle_pd(const double * gmx_restrict p1,
 248                                   __m128d * gmx_restrict x, __m128d * gmx_restrict y, __m128d * gmx_restrict z)
 249 {
 250     *x            = _mm_load_sd(p1);
 251     *y            = _mm_load_sd(p1+1);
 252     *z            = _mm_load_sd(p1+2);
 253 }
 254
 255 static gmx_inline void
 256 gmx_mm_load_3rvec_1ptr_swizzle_pd(const double * gmx_restrict p1,
 257                                   __m128d * gmx_restrict x1, __m128d * gmx_restrict y1, __m128d * gmx_restrict z1,
 258                                   __m128d * gmx_restrict x2, __m128d * gmx_restrict y2, __m128d * gmx_restrict z2,
 259                                   __m128d * gmx_restrict x3, __m128d * gmx_restrict y3, __m128d * gmx_restrict z3)
 260 {
 261     *x1            = _mm_load_sd(p1);
 262     *y1            = _mm_load_sd(p1+1);
 263     *z1            = _mm_load_sd(p1+2);
 264     *x2            = _mm_load_sd(p1+3);
 265     *y2            = _mm_load_sd(p1+4);
 266     *z2            = _mm_load_sd(p1+5);
 267     *x3            = _mm_load_sd(p1+6);
 268     *y3            = _mm_load_sd(p1+7);
 269     *z3            = _mm_load_sd(p1+8);
 270 }
 271
 272 static gmx_inline void
 273 gmx_mm_load_4rvec_1ptr_swizzle_pd(const double * gmx_restrict p1,
 274                                   __m128d * gmx_restrict x1, __m128d * gmx_restrict y1, __m128d * gmx_restrict z1,
 275                                   __m128d * gmx_restrict x2, __m128d * gmx_restrict y2, __m128d * gmx_restrict z2,
 276                                   __m128d * gmx_restrict x3, __m128d * gmx_restrict y3, __m128d * gmx_restrict z3,
 277                                   __m128d * gmx_restrict x4, __m128d * gmx_restrict y4, __m128d * gmx_restrict z4)
 278 {
 279     *x1            = _mm_load_sd(p1);
 280     *y1            = _mm_load_sd(p1+1);
 281     *z1            = _mm_load_sd(p1+2);
 282     *x2            = _mm_load_sd(p1+3);
 283     *y2            = _mm_load_sd(p1+4);
 284     *z2            = _mm_load_sd(p1+5);
 285     *x3            = _mm_load_sd(p1+6);
 286     *y3            = _mm_load_sd(p1+7);
 287     *z3            = _mm_load_sd(p1+8);
 288     *x4            = _mm_load_sd(p1+9);
 289     *y4            = _mm_load_sd(p1+10);
 290     *z4            = _mm_load_sd(p1+11);
 291 }
 292
 293
 294 static gmx_inline void
 295 gmx_mm_load_1rvec_2ptr_swizzle_pd(const double * gmx_restrict ptrA,
 296                                   const double * gmx_restrict ptrB,
 297                                   __m128d * gmx_restrict x1, __m128d * gmx_restrict y1, __m128d * gmx_restrict z1)
 298 {
 299     __m128d t1,t2,t3,t4;
 300     t1           = _mm_loadu_pd(ptrA);
 301     t2           = _mm_loadu_pd(ptrB);
 302     t3           = _mm_load_sd(ptrA+2);
 303     t4           = _mm_load_sd(ptrB+2);
 304     GMX_MM_TRANSPOSE2_PD(t1,t2);
 305     *x1          = t1;
 306     *y1          = t2;
 307     *z1          = _mm_unpacklo_pd(t3,t4);
 308 }
 309
 310
 311 static gmx_inline void
 312 gmx_mm_load_3rvec_2ptr_swizzle_pd(const double * gmx_restrict ptrA, const double * gmx_restrict ptrB,
 313                                   __m128d * gmx_restrict x1, __m128d * gmx_restrict y1, __m128d * gmx_restrict z1,
 314                                   __m128d * gmx_restrict x2, __m128d * gmx_restrict y2, __m128d * gmx_restrict z2,
 315                                   __m128d * gmx_restrict x3, __m128d * gmx_restrict y3, __m128d * gmx_restrict z3)
 316 {
 317     __m128d t1,t2,t3,t4,t5,t6,t7,t8,t9,t10;
 318     t1           = _mm_loadu_pd(ptrA);
 319     t2           = _mm_loadu_pd(ptrB);
 320     t3           = _mm_loadu_pd(ptrA+2);
 321     t4           = _mm_loadu_pd(ptrB+2);
 322     t5           = _mm_loadu_pd(ptrA+4);
 323     t6           = _mm_loadu_pd(ptrB+4);
 324     t7           = _mm_loadu_pd(ptrA+6);
 325     t8           = _mm_loadu_pd(ptrB+6);
 326     t9           = _mm_load_sd(ptrA+8);
 327     t10          = _mm_load_sd(ptrB+8);
 328     GMX_MM_TRANSPOSE2_PD(t1,t2);
 329     GMX_MM_TRANSPOSE2_PD(t3,t4);
 330     GMX_MM_TRANSPOSE2_PD(t5,t6);
 331     GMX_MM_TRANSPOSE2_PD(t7,t8);
 332     *x1          = t1;
 333     *y1          = t2;
 334     *z1          = t3;
 335     *x2          = t4;
 336     *y2          = t5;
 337     *z2          = t6;
 338     *x3          = t7;
 339     *y3          = t8;
 340     *z3          = _mm_unpacklo_pd(t9,t10);
 341 }
 342
 343
 344 static gmx_inline void
 345 gmx_mm_load_4rvec_2ptr_swizzle_pd(const double * gmx_restrict ptrA, const double * gmx_restrict ptrB,
 346                                   __m128d * gmx_restrict x1, __m128d * gmx_restrict y1, __m128d * gmx_restrict z1,
 347                                   __m128d * gmx_restrict x2, __m128d * gmx_restrict y2, __m128d * gmx_restrict z2,
 348                                   __m128d * gmx_restrict x3, __m128d * gmx_restrict y3, __m128d * gmx_restrict z3,
 349                                   __m128d * gmx_restrict x4, __m128d * gmx_restrict y4, __m128d * gmx_restrict z4)
 350 {
 351     __m128d t1,t2,t3,t4,t5,t6;
 352     t1           = _mm_loadu_pd(ptrA);
 353     t2           = _mm_loadu_pd(ptrB);
 354     t3           = _mm_loadu_pd(ptrA+2);
 355     t4           = _mm_loadu_pd(ptrB+2);
 356     t5           = _mm_loadu_pd(ptrA+4);
 357     t6           = _mm_loadu_pd(ptrB+4);
 358     GMX_MM_TRANSPOSE2_PD(t1,t2);
 359     GMX_MM_TRANSPOSE2_PD(t3,t4);
 360     GMX_MM_TRANSPOSE2_PD(t5,t6);
 361     *x1          = t1;
 362     *y1          = t2;
 363     *z1          = t3;
 364     *x2          = t4;
 365     *y2          = t5;
 366     *z2          = t6;
 367     t1           = _mm_loadu_pd(ptrA+6);
 368     t2           = _mm_loadu_pd(ptrB+6);
 369     t3           = _mm_loadu_pd(ptrA+8);
 370     t4           = _mm_loadu_pd(ptrB+8);
 371     t5           = _mm_loadu_pd(ptrA+10);
 372     t6           = _mm_loadu_pd(ptrB+10);
 373     GMX_MM_TRANSPOSE2_PD(t1,t2);
 374     GMX_MM_TRANSPOSE2_PD(t3,t4);
 375     GMX_MM_TRANSPOSE2_PD(t5,t6);
 376     *x3          = t1;
 377     *y3          = t2;
 378     *z3          = t3;
 379     *x4          = t4;
 380     *y4          = t5;
 381     *z4          = t6;
 382 }
 383
 384
 385 /* Routines to decrement rvec in memory, typically use for j particle force updates */
 386 static gmx_inline void
 387 gmx_mm_decrement_1rvec_1ptr_noswizzle_pd(double * gmx_restrict ptrA,
 388         __m128d xy, __m128d z)
 389 {
 390     __m128d t1,t2;
 391
 392     t1 = _mm_loadu_pd(ptrA);
 393     t2 = _mm_load_sd(ptrA+2);
 394
 395     t1 = _mm_sub_pd(t1,xy);
 396     t2 = _mm_sub_sd(t2,z);
 397
 398     _mm_storeu_pd(ptrA,t1);
 399     _mm_store_sd(ptrA+2,t2);
 400 }
 401
 402
 403 static gmx_inline void
 404 gmx_mm_decrement_1rvec_1ptr_swizzle_pd(double * gmx_restrict ptrA,
 405                                        __m128d x1, __m128d y1, __m128d z1)
 406 {
 407     __m128d t1,t2,t3;
 408
 409     t1           = _mm_load_sd(ptrA);
 410     t2           = _mm_load_sd(ptrA+1);
 411     t3           = _mm_load_sd(ptrA+2);
 412
 413     t1           = _mm_sub_sd(t1,x1);
 414     t2           = _mm_sub_sd(t2,y1);
 415     t3           = _mm_sub_sd(t3,z1);
 416     _mm_store_sd(ptrA,t1);
 417     _mm_store_sd(ptrA+1,t2);
 418     _mm_store_sd(ptrA+2,t3);
 419 }
 420
 421
 422 #if defined (_MSC_VER) && defined(_M_IX86)
 423 /* Macro work-around since 32-bit MSVC cannot handle >3 xmm/ymm parameters */
 424 #define gmx_mm_decrement_3rvec_1ptr_swizzle_pd(ptrA,_x1,_y1,_z1,_x2,_y2,_z2,_x3,_y3,_z3) \
 425 {\
 426     __m128d _t1,_t2,_t3,_t4,_t5;\
 427     _t1          = _mm_loadu_pd(ptrA);\
 428     _t2          = _mm_loadu_pd(ptrA+2);\
 429     _t3          = _mm_loadu_pd(ptrA+4);\
 430     _t4          = _mm_loadu_pd(ptrA+6);\
 431     _t5          = _mm_load_sd(ptrA+8);\
 432     _x1          = _mm_unpacklo_pd(_x1,_y1);\
 433     _z1          = _mm_unpacklo_pd(_z1,_x2);\
 434     _y2          = _mm_unpacklo_pd(_y2,_z2);\
 435     _x3          = _mm_unpacklo_pd(_x3,_y3);\
 436     _t1          = _mm_sub_pd(_t1,_x1);\
 437     _t2          = _mm_sub_pd(_t2,_z1);\
 438     _t3          = _mm_sub_pd(_t3,_y2);\
 439     _t4          = _mm_sub_pd(_t4,_x3);\
 440     _t5          = _mm_sub_sd(_t5,_z3);\
 441     _mm_storeu_pd(ptrA,_t1);\
 442     _mm_storeu_pd(ptrA+2,_t2);\
 443     _mm_storeu_pd(ptrA+4,_t3);\
 444     _mm_storeu_pd(ptrA+6,_t4);\
 445     _mm_store_sd(ptrA+8,_t5);\
 446 }
 447 #else
 448 /* Real function for sane compilers */
 449 static gmx_inline void
 450 gmx_mm_decrement_3rvec_1ptr_swizzle_pd(double * gmx_restrict ptrA,
 451                                        __m128d x1, __m128d y1, __m128d z1,
 452                                        __m128d x2, __m128d y2, __m128d z2,
 453                                        __m128d x3, __m128d y3, __m128d z3)
 454 {
 455     __m128d t1,t2,t3,t4,t5;
 456
 457     t1          = _mm_loadu_pd(ptrA);
 458     t2          = _mm_loadu_pd(ptrA+2);
 459     t3          = _mm_loadu_pd(ptrA+4);
 460     t4          = _mm_loadu_pd(ptrA+6);
 461     t5          = _mm_load_sd(ptrA+8);
 462
 463     x1          = _mm_unpacklo_pd(x1,y1);
 464     z1          = _mm_unpacklo_pd(z1,x2);
 465     y2          = _mm_unpacklo_pd(y2,z2);
 466     x3          = _mm_unpacklo_pd(x3,y3);
 467     /* nothing to be done for z3 */
 468
 469     t1          = _mm_sub_pd(t1,x1);
 470     t2          = _mm_sub_pd(t2,z1);
 471     t3          = _mm_sub_pd(t3,y2);
 472     t4          = _mm_sub_pd(t4,x3);
 473     t5          = _mm_sub_sd(t5,z3);
 474     _mm_storeu_pd(ptrA,t1);
 475     _mm_storeu_pd(ptrA+2,t2);
 476     _mm_storeu_pd(ptrA+4,t3);
 477     _mm_storeu_pd(ptrA+6,t4);
 478     _mm_store_sd(ptrA+8,t5);
 479 }
 480 #endif
 481
 482
 483 #if defined (_MSC_VER) && defined(_M_IX86)
 484 /* Macro work-around since 32-bit MSVC cannot handle >3 xmm/ymm parameters */
 485 #define gmx_mm_decrement_4rvec_1ptr_swizzle_pd(ptrA,_x1,_y1,_z1,_x2,_y2,_z2,_x3,_y3,_z3,_x4,_y4,_z4) \
 486 {\
 487     __m128d _t1,_t2,_t3,_t4,_t5,_t6;\
 488     _t1          = _mm_loadu_pd(ptrA);\
 489     _t2          = _mm_loadu_pd(ptrA+2);\
 490     _t3          = _mm_loadu_pd(ptrA+4);\
 491     _t4          = _mm_loadu_pd(ptrA+6);\
 492     _t5          = _mm_loadu_pd(ptrA+8);\
 493     _t6          = _mm_loadu_pd(ptrA+10);\
 494     _x1          = _mm_unpacklo_pd(_x1,_y1);\
 495     _z1          = _mm_unpacklo_pd(_z1,_x2);\
 496     _y2          = _mm_unpacklo_pd(_y2,_z2);\
 497     _x3          = _mm_unpacklo_pd(_x3,_y3);\
 498     _z3          = _mm_unpacklo_pd(_z3,_x4);\
 499     _y4          = _mm_unpacklo_pd(_y4,_z4);\
 500     _mm_storeu_pd(ptrA,    _mm_sub_pd( _t1,_x1 ));\
 501     _mm_storeu_pd(ptrA+2,  _mm_sub_pd( _t2,_z1 ));\
 502     _mm_storeu_pd(ptrA+4,  _mm_sub_pd( _t3,_y2 ));\
 503     _mm_storeu_pd(ptrA+6,  _mm_sub_pd( _t4,_x3 ));\
 504     _mm_storeu_pd(ptrA+8,  _mm_sub_pd( _t5,_z3 ));\
 505     _mm_storeu_pd(ptrA+10, _mm_sub_pd( _t6,_y4 ));\
 506 }
 507 #else
 508 /* Real function for sane compilers */
 509 static gmx_inline void
 510 gmx_mm_decrement_4rvec_1ptr_swizzle_pd(double * gmx_restrict ptrA,
 511                                        __m128d x1, __m128d y1, __m128d z1,
 512                                        __m128d x2, __m128d y2, __m128d z2,
 513                                        __m128d x3, __m128d y3, __m128d z3,
 514                                        __m128d x4, __m128d y4, __m128d z4)
 515 {
 516     __m128d t1,t2,t3,t4,t5,t6;
 517
 518     t1          = _mm_loadu_pd(ptrA);
 519     t2          = _mm_loadu_pd(ptrA+2);
 520     t3          = _mm_loadu_pd(ptrA+4);
 521     t4          = _mm_loadu_pd(ptrA+6);
 522     t5          = _mm_loadu_pd(ptrA+8);
 523     t6          = _mm_loadu_pd(ptrA+10);
 524
 525     x1          = _mm_unpacklo_pd(x1,y1);
 526     z1          = _mm_unpacklo_pd(z1,x2);
 527     y2          = _mm_unpacklo_pd(y2,z2);
 528     x3          = _mm_unpacklo_pd(x3,y3);
 529     z3          = _mm_unpacklo_pd(z3,x4);
 530     y4          = _mm_unpacklo_pd(y4,z4);
 531
 532     _mm_storeu_pd(ptrA,    _mm_sub_pd( t1,x1 ));
 533     _mm_storeu_pd(ptrA+2,  _mm_sub_pd( t2,z1 ));
 534     _mm_storeu_pd(ptrA+4,  _mm_sub_pd( t3,y2 ));
 535     _mm_storeu_pd(ptrA+6,  _mm_sub_pd( t4,x3 ));
 536     _mm_storeu_pd(ptrA+8,  _mm_sub_pd( t5,z3 ));
 537     _mm_storeu_pd(ptrA+10, _mm_sub_pd( t6,y4 ));
 538 }
 539 #endif
 540
 541
 542 static gmx_inline void
 543 gmx_mm_decrement_1rvec_2ptr_swizzle_pd(double * gmx_restrict ptrA, double * gmx_restrict ptrB,
 544                                        __m128d x1, __m128d y1, __m128d z1)
 545 {
 546     __m128d t1,t2,t3,t4,t5,t6,t7;
 547
 548     t1          = _mm_loadu_pd(ptrA);
 549     t2          = _mm_load_sd(ptrA+2);
 550     t3          = _mm_loadu_pd(ptrB);
 551     t4          = _mm_load_sd(ptrB+2);
 552
 553     t5          = _mm_unpacklo_pd(x1,y1);
 554     t6          = _mm_unpackhi_pd(x1,y1);
 555     t7          = _mm_unpackhi_pd(z1,z1);
 556
 557     t1          = _mm_sub_pd(t1,t5);
 558     t2          = _mm_sub_sd(t2,z1);
 559
 560     t3          = _mm_sub_pd(t3,t6);
 561     t4          = _mm_sub_sd(t4,t7);
 562
 563     _mm_storeu_pd(ptrA,t1);
 564     _mm_store_sd(ptrA+2,t2);
 565     _mm_storeu_pd(ptrB,t3);
 566     _mm_store_sd(ptrB+2,t4);
 567 }
 568
 569 #if defined (_MSC_VER) && defined(_M_IX86)
 570 /* Macro work-around since 32-bit MSVC cannot handle >3 xmm/ymm parameters */
 571 #define gmx_mm_decrement_3rvec_2ptr_swizzle_pd(ptrA,ptrB,_x1,_y1,_z1,_x2,_y2,_z2,_x3,_y3,_z3) \
 572 {\
 573     __m128d _t1,_t2,_t3,_t4,_t5,_t6,_t7,_t8,_t9,_t10;\
 574     __m128d _tA,_tB,_tC,_tD,_tE,_tF,_tG,_tH,_tI;\
 575     _t1          = _mm_loadu_pd(ptrA);\
 576     _t2          = _mm_loadu_pd(ptrA+2);\
 577     _t3          = _mm_loadu_pd(ptrA+4);\
 578     _t4          = _mm_loadu_pd(ptrA+6);\
 579     _t5          = _mm_load_sd(ptrA+8);\
 580     _t6          = _mm_loadu_pd(ptrB);\
 581     _t7          = _mm_loadu_pd(ptrB+2);\
 582     _t8          = _mm_loadu_pd(ptrB+4);\
 583     _t9          = _mm_loadu_pd(ptrB+6);\
 584     _t10         = _mm_load_sd(ptrB+8);\
 585     _tA          = _mm_unpacklo_pd(_x1,_y1);\
 586     _tB          = _mm_unpackhi_pd(_x1,_y1);\
 587     _tC          = _mm_unpacklo_pd(_z1,_x2);\
 588     _tD          = _mm_unpackhi_pd(_z1,_x2);\
 589     _tE          = _mm_unpacklo_pd(_y2,_z2);\
 590     _tF          = _mm_unpackhi_pd(_y2,_z2);\
 591     _tG          = _mm_unpacklo_pd(_x3,_y3);\
 592     _tH          = _mm_unpackhi_pd(_x3,_y3);\
 593     _tI          = _mm_unpackhi_pd(_z3,_z3);\
 594     _t1          = _mm_sub_pd(_t1,_tA);\
 595     _t2          = _mm_sub_pd(_t2,_tC);\
 596     _t3          = _mm_sub_pd(_t3,_tE);\
 597     _t4          = _mm_sub_pd(_t4,_tG);\
 598     _t5          = _mm_sub_sd(_t5,_z3);\
 599     _t6          = _mm_sub_pd(_t6,_tB);\
 600     _t7          = _mm_sub_pd(_t7,_tD);\
 601     _t8          = _mm_sub_pd(_t8,_tF);\
 602     _t9          = _mm_sub_pd(_t9,_tH);\
 603     _t10         = _mm_sub_sd(_t10,_tI);\
 604     _mm_storeu_pd(ptrA,_t1);\
 605     _mm_storeu_pd(ptrA+2,_t2);\
 606     _mm_storeu_pd(ptrA+4,_t3);\
 607     _mm_storeu_pd(ptrA+6,_t4);\
 608     _mm_store_sd(ptrA+8,_t5);\
 609     _mm_storeu_pd(ptrB,_t6);\
 610     _mm_storeu_pd(ptrB+2,_t7);\
 611     _mm_storeu_pd(ptrB+4,_t8);\
 612     _mm_storeu_pd(ptrB+6,_t9);\
 613     _mm_store_sd(ptrB+8,_t10);\
 614 }
 615 #else
 616 /* Real function for sane compilers */
 617 static gmx_inline void
 618 gmx_mm_decrement_3rvec_2ptr_swizzle_pd(double * gmx_restrict ptrA, double * gmx_restrict ptrB,
 619                                        __m128d x1, __m128d y1, __m128d z1,
 620                                        __m128d x2, __m128d y2, __m128d z2,
 621                                        __m128d x3, __m128d y3, __m128d z3)
 622 {
 623     __m128d t1,t2,t3,t4,t5,t6,t7,t8,t9,t10;
 624     __m128d tA,tB,tC,tD,tE,tF,tG,tH,tI;
 625
 626     t1          = _mm_loadu_pd(ptrA);
 627     t2          = _mm_loadu_pd(ptrA+2);
 628     t3          = _mm_loadu_pd(ptrA+4);
 629     t4          = _mm_loadu_pd(ptrA+6);
 630     t5          = _mm_load_sd(ptrA+8);
 631     t6          = _mm_loadu_pd(ptrB);
 632     t7          = _mm_loadu_pd(ptrB+2);
 633     t8          = _mm_loadu_pd(ptrB+4);
 634     t9          = _mm_loadu_pd(ptrB+6);
 635     t10         = _mm_load_sd(ptrB+8);
 636
 637     tA          = _mm_unpacklo_pd(x1,y1);
 638     tB          = _mm_unpackhi_pd(x1,y1);
 639     tC          = _mm_unpacklo_pd(z1,x2);
 640     tD          = _mm_unpackhi_pd(z1,x2);
 641     tE          = _mm_unpacklo_pd(y2,z2);
 642     tF          = _mm_unpackhi_pd(y2,z2);
 643     tG          = _mm_unpacklo_pd(x3,y3);
 644     tH          = _mm_unpackhi_pd(x3,y3);
 645     tI          = _mm_unpackhi_pd(z3,z3);
 646
 647     t1          = _mm_sub_pd(t1,tA);
 648     t2          = _mm_sub_pd(t2,tC);
 649     t3          = _mm_sub_pd(t3,tE);
 650     t4          = _mm_sub_pd(t4,tG);
 651     t5          = _mm_sub_sd(t5,z3);
 652
 653     t6          = _mm_sub_pd(t6,tB);
 654     t7          = _mm_sub_pd(t7,tD);
 655     t8          = _mm_sub_pd(t8,tF);
 656     t9          = _mm_sub_pd(t9,tH);
 657     t10         = _mm_sub_sd(t10,tI);
 658
 659     _mm_storeu_pd(ptrA,t1);
 660     _mm_storeu_pd(ptrA+2,t2);
 661     _mm_storeu_pd(ptrA+4,t3);
 662     _mm_storeu_pd(ptrA+6,t4);
 663     _mm_store_sd(ptrA+8,t5);
 664     _mm_storeu_pd(ptrB,t6);
 665     _mm_storeu_pd(ptrB+2,t7);
 666     _mm_storeu_pd(ptrB+4,t8);
 667     _mm_storeu_pd(ptrB+6,t9);
 668     _mm_store_sd(ptrB+8,t10);
 669 }
 670 #endif
 671
 672
 673 #if defined (_MSC_VER) && defined(_M_IX86)
 674 /* Macro work-around since 32-bit MSVC cannot handle >3 xmm/ymm parameters */
 675 #define gmx_mm_decrement_4rvec_2ptr_swizzle_pd(ptrA,ptrB,_x1,_y1,_z1,_x2,_y2,_z2,_x3,_y3,_z3,_x4,_y4,_z4) \
 676 {\
 677     __m128d _t1,_t2,_t3,_t4,_t5,_t6,_t7,_t8,_t9,_t10,_t11,_t12;\
 678     __m128d _tA,_tB,_tC,_tD,_tE,_tF,_tG,_tH,_tI,_tJ,_tK,_tL;\
 679     _t1          = _mm_loadu_pd(ptrA);\
 680     _t2          = _mm_loadu_pd(ptrA+2);\
 681     _t3          = _mm_loadu_pd(ptrA+4);\
 682     _t4          = _mm_loadu_pd(ptrA+6);\
 683     _t5          = _mm_loadu_pd(ptrA+8);\
 684     _t6          = _mm_loadu_pd(ptrA+10);\
 685     _t7          = _mm_loadu_pd(ptrB);\
 686     _t8          = _mm_loadu_pd(ptrB+2);\
 687     _t9          = _mm_loadu_pd(ptrB+4);\
 688     _t10         = _mm_loadu_pd(ptrB+6);\
 689     _t11         = _mm_loadu_pd(ptrB+8);\
 690     _t12         = _mm_loadu_pd(ptrB+10);\
 691     _tA          = _mm_unpacklo_pd(_x1,_y1);\
 692     _tB          = _mm_unpackhi_pd(_x1,_y1);\
 693     _tC          = _mm_unpacklo_pd(_z1,_x2);\
 694     _tD          = _mm_unpackhi_pd(_z1,_x2);\
 695     _tE          = _mm_unpacklo_pd(_y2,_z2);\
 696     _tF          = _mm_unpackhi_pd(_y2,_z2);\
 697     _tG          = _mm_unpacklo_pd(_x3,_y3);\
 698     _tH          = _mm_unpackhi_pd(_x3,_y3);\
 699     _tI          = _mm_unpacklo_pd(_z3,_x4);\
 700     _tJ          = _mm_unpackhi_pd(_z3,_x4);\
 701     _tK          = _mm_unpacklo_pd(_y4,_z4);\
 702     _tL          = _mm_unpackhi_pd(_y4,_z4);\
 703     _t1          = _mm_sub_pd(_t1,_tA);\
 704     _t2          = _mm_sub_pd(_t2,_tC);\
 705     _t3          = _mm_sub_pd(_t3,_tE);\
 706     _t4          = _mm_sub_pd(_t4,_tG);\
 707     _t5          = _mm_sub_pd(_t5,_tI);\
 708     _t6          = _mm_sub_pd(_t6,_tK);\
 709     _t7          = _mm_sub_pd(_t7,_tB);\
 710     _t8          = _mm_sub_pd(_t8,_tD);\
 711     _t9          = _mm_sub_pd(_t9,_tF);\
 712     _t10         = _mm_sub_pd(_t10,_tH);\
 713     _t11         = _mm_sub_pd(_t11,_tJ);\
 714     _t12         = _mm_sub_pd(_t12,_tL);\
 715     _mm_storeu_pd(ptrA,  _t1);\
 716     _mm_storeu_pd(ptrA+2,_t2);\
 717     _mm_storeu_pd(ptrA+4,_t3);\
 718     _mm_storeu_pd(ptrA+6,_t4);\
 719     _mm_storeu_pd(ptrA+8,_t5);\
 720     _mm_storeu_pd(ptrA+10,_t6);\
 721     _mm_storeu_pd(ptrB,  _t7);\
 722     _mm_storeu_pd(ptrB+2,_t8);\
 723     _mm_storeu_pd(ptrB+4,_t9);\
 724     _mm_storeu_pd(ptrB+6,_t10);\
 725     _mm_storeu_pd(ptrB+8,_t11);\
 726     _mm_storeu_pd(ptrB+10,_t12);\
 727 }
 728 #else
 729 /* Real function for sane compilers */
 730 static gmx_inline void
 731 gmx_mm_decrement_4rvec_2ptr_swizzle_pd(double * gmx_restrict ptrA, double * gmx_restrict ptrB,
 732                                        __m128d x1, __m128d y1, __m128d z1,
 733                                        __m128d x2, __m128d y2, __m128d z2,
 734                                        __m128d x3, __m128d y3, __m128d z3,
 735                                        __m128d x4, __m128d y4, __m128d z4)
 736 {
 737     __m128d t1,t2,t3,t4,t5,t6,t7,t8,t9,t10,t11,t12;
 738     __m128d tA,tB,tC,tD,tE,tF,tG,tH,tI,tJ,tK,tL;
 739
 740     t1          = _mm_loadu_pd(ptrA);
 741     t2          = _mm_loadu_pd(ptrA+2);
 742     t3          = _mm_loadu_pd(ptrA+4);
 743     t4          = _mm_loadu_pd(ptrA+6);
 744     t5          = _mm_loadu_pd(ptrA+8);
 745     t6          = _mm_loadu_pd(ptrA+10);
 746     t7          = _mm_loadu_pd(ptrB);
 747     t8          = _mm_loadu_pd(ptrB+2);
 748     t9          = _mm_loadu_pd(ptrB+4);
 749     t10         = _mm_loadu_pd(ptrB+6);
 750     t11         = _mm_loadu_pd(ptrB+8);
 751     t12         = _mm_loadu_pd(ptrB+10);
 752
 753     tA          = _mm_unpacklo_pd(x1,y1);
 754     tB          = _mm_unpackhi_pd(x1,y1);
 755     tC          = _mm_unpacklo_pd(z1,x2);
 756     tD          = _mm_unpackhi_pd(z1,x2);
 757     tE          = _mm_unpacklo_pd(y2,z2);
 758     tF          = _mm_unpackhi_pd(y2,z2);
 759     tG          = _mm_unpacklo_pd(x3,y3);
 760     tH          = _mm_unpackhi_pd(x3,y3);
 761     tI          = _mm_unpacklo_pd(z3,x4);
 762     tJ          = _mm_unpackhi_pd(z3,x4);
 763     tK          = _mm_unpacklo_pd(y4,z4);
 764     tL          = _mm_unpackhi_pd(y4,z4);
 765
 766     t1          = _mm_sub_pd(t1,tA);
 767     t2          = _mm_sub_pd(t2,tC);
 768     t3          = _mm_sub_pd(t3,tE);
 769     t4          = _mm_sub_pd(t4,tG);
 770     t5          = _mm_sub_pd(t5,tI);
 771     t6          = _mm_sub_pd(t6,tK);
 772
 773     t7          = _mm_sub_pd(t7,tB);
 774     t8          = _mm_sub_pd(t8,tD);
 775     t9          = _mm_sub_pd(t9,tF);
 776     t10         = _mm_sub_pd(t10,tH);
 777     t11         = _mm_sub_pd(t11,tJ);
 778     t12         = _mm_sub_pd(t12,tL);
 779
 780     _mm_storeu_pd(ptrA,  t1);
 781     _mm_storeu_pd(ptrA+2,t2);
 782     _mm_storeu_pd(ptrA+4,t3);
 783     _mm_storeu_pd(ptrA+6,t4);
 784     _mm_storeu_pd(ptrA+8,t5);
 785     _mm_storeu_pd(ptrA+10,t6);
 786     _mm_storeu_pd(ptrB,  t7);
 787     _mm_storeu_pd(ptrB+2,t8);
 788     _mm_storeu_pd(ptrB+4,t9);
 789     _mm_storeu_pd(ptrB+6,t10);
 790     _mm_storeu_pd(ptrB+8,t11);
 791     _mm_storeu_pd(ptrB+10,t12);
 792 }
 793 #endif
 794
 795
 796
 797
 798
 799 static gmx_inline void
 800 gmx_mm_update_iforce_1atom_swizzle_pd(__m128d fix1, __m128d fiy1, __m128d fiz1,
 801                                       double * gmx_restrict fptr,
 802                                       double * gmx_restrict fshiftptr)
 803 {
 804     __m128d t1,t2,t3;
 805
 806     /* transpose data */
 807     t1 = fix1;
 808     fix1 = _mm_unpacklo_pd(fix1,fiy1); /* y0 x0 */
 809     fiy1 = _mm_unpackhi_pd(t1,fiy1);   /* y1 x1 */
 810
 811     fix1 = _mm_add_pd(fix1,fiy1);
 812     fiz1 = _mm_add_sd( fiz1, _mm_unpackhi_pd(fiz1,fiz1 ));
 813
 814     _mm_storeu_pd( fptr, _mm_add_pd( _mm_loadu_pd(fptr), fix1 ));
 815     _mm_store_sd( fptr+2, _mm_add_sd( _mm_load_sd(fptr+2), fiz1 ));
 816
 817     _mm_storeu_pd( fshiftptr, _mm_add_pd( _mm_loadu_pd(fshiftptr), fix1 ));
 818     _mm_store_sd( fshiftptr+2, _mm_add_sd( _mm_load_sd(fshiftptr+2), fiz1 ));
 819 }
 820
 821 #if defined (_MSC_VER) && defined(_M_IX86)
 822 /* Macro work-around since 32-bit MSVC cannot handle >3 xmm/ymm parameters */
 823 #define gmx_mm_update_iforce_3atom_swizzle_pd(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3, \
 824                                               fptr,fshiftptr) \
 825 {\
 826     __m128d _t1,_t2;\
 827     GMX_MM_TRANSPOSE2_PD(fix1,fiy1);\
 828     GMX_MM_TRANSPOSE2_PD(fiz1,fix2);\
 829     GMX_MM_TRANSPOSE2_PD(fiy2,fiz2);\
 830     _t1 = fix3;\
 831     fix3 = _mm_unpacklo_pd(fix3,fiy3);\
 832     fiy3 = _mm_unpackhi_pd(_t1,fiy3);\
 833     fix1 = _mm_add_pd(fix1,fiy1);\
 834     fiz1 = _mm_add_pd(fiz1,fix2);\
 835     fiy2 = _mm_add_pd(fiy2,fiz2);\
 836     fix3 = _mm_add_pd(fix3,fiy3);\
 837     fiz3 = _mm_add_sd( fiz3, _mm_unpackhi_pd(fiz3,fiz3));\
 838     _mm_storeu_pd( fptr, _mm_add_pd( _mm_loadu_pd(fptr), fix1 ));\
 839     _mm_storeu_pd( fptr+2, _mm_add_pd( _mm_loadu_pd(fptr+2), fiz1 ));\
 840     _mm_storeu_pd( fptr+4, _mm_add_pd( _mm_loadu_pd(fptr+4), fiy2 ));\
 841     _mm_storeu_pd( fptr+6, _mm_add_pd( _mm_loadu_pd(fptr+6), fix3 ));\
 842     _mm_store_sd( fptr+8, _mm_add_sd( _mm_load_sd(fptr+8), fiz3 ));\
 843     fix1 = _mm_add_pd(fix1,fix3);\
 844     _t1   = _mm_shuffle_pd(fiz1,fiy2,_MM_SHUFFLE2(0,1));\
 845     fix1 = _mm_add_pd(fix1,_t1);\
 846     _t2   = _mm_shuffle_pd(fiy2,fiy2,_MM_SHUFFLE2(1,1));\
 847     fiz1 = _mm_add_sd(fiz1,fiz3);\
 848     fiz1 = _mm_add_sd(fiz1,_t2);\
 849     _mm_storeu_pd( fshiftptr, _mm_add_pd( _mm_loadu_pd(fshiftptr), fix1 ));\
 850     _mm_store_sd( fshiftptr+2, _mm_add_sd( _mm_load_sd(fshiftptr+2), fiz1 ));\
 851 }
 852 #else
 853 /* Real function for sane compilers */
 854 static gmx_inline void
 855 gmx_mm_update_iforce_3atom_swizzle_pd(__m128d fix1, __m128d fiy1, __m128d fiz1,
 856                                       __m128d fix2, __m128d fiy2, __m128d fiz2,
 857                                       __m128d fix3, __m128d fiy3, __m128d fiz3,
 858                                       double * gmx_restrict fptr,
 859                                       double * gmx_restrict fshiftptr)
 860 {
 861     __m128d t1,t2;
 862
 863     /* transpose data */
 864     GMX_MM_TRANSPOSE2_PD(fix1,fiy1);
 865     GMX_MM_TRANSPOSE2_PD(fiz1,fix2);
 866     GMX_MM_TRANSPOSE2_PD(fiy2,fiz2);
 867     t1 = fix3;
 868     fix3 = _mm_unpacklo_pd(fix3,fiy3); /* y0 x0 */
 869     fiy3 = _mm_unpackhi_pd(t1,fiy3);   /* y1 x1 */
 870
 871     fix1 = _mm_add_pd(fix1,fiy1);
 872     fiz1 = _mm_add_pd(fiz1,fix2);
 873     fiy2 = _mm_add_pd(fiy2,fiz2);
 874
 875     fix3 = _mm_add_pd(fix3,fiy3);
 876     fiz3 = _mm_add_sd( fiz3, _mm_unpackhi_pd(fiz3,fiz3));
 877
 878     _mm_storeu_pd( fptr, _mm_add_pd( _mm_loadu_pd(fptr), fix1 ));
 879     _mm_storeu_pd( fptr+2, _mm_add_pd( _mm_loadu_pd(fptr+2), fiz1 ));
 880     _mm_storeu_pd( fptr+4, _mm_add_pd( _mm_loadu_pd(fptr+4), fiy2 ));
 881     _mm_storeu_pd( fptr+6, _mm_add_pd( _mm_loadu_pd(fptr+6), fix3 ));
 882     _mm_store_sd( fptr+8, _mm_add_sd( _mm_load_sd(fptr+8), fiz3 ));
 883
 884     fix1 = _mm_add_pd(fix1,fix3);
 885     t1   = _mm_shuffle_pd(fiz1,fiy2,_MM_SHUFFLE2(0,1));
 886     fix1 = _mm_add_pd(fix1,t1); /* x and y sums */
 887
 888     t2   = _mm_shuffle_pd(fiy2,fiy2,_MM_SHUFFLE2(1,1));
 889     fiz1 = _mm_add_sd(fiz1,fiz3);
 890     fiz1 = _mm_add_sd(fiz1,t2); /* z sum */
 891
 892     _mm_storeu_pd( fshiftptr, _mm_add_pd( _mm_loadu_pd(fshiftptr), fix1 ));
 893     _mm_store_sd( fshiftptr+2, _mm_add_sd( _mm_load_sd(fshiftptr+2), fiz1 ));
 894 }
 895 #endif
 896
 897
 898 #if defined (_MSC_VER) && defined(_M_IX86)
 899 /* Macro work-around since 32-bit MSVC cannot handle >3 xmm/ymm parameters */
 900 #define gmx_mm_update_iforce_4atom_swizzle_pd(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,fix4,fiy4,fiz4, \
 901                                               fptr,fshiftptr) \
 902 {\
 903     __m128d _t1,_t2;\
 904     GMX_MM_TRANSPOSE2_PD(fix1,fiy1);\
 905     GMX_MM_TRANSPOSE2_PD(fiz1,fix2);\
 906     GMX_MM_TRANSPOSE2_PD(fiy2,fiz2);\
 907     GMX_MM_TRANSPOSE2_PD(fix3,fiy3);\
 908     GMX_MM_TRANSPOSE2_PD(fiz3,fix4);\
 909     GMX_MM_TRANSPOSE2_PD(fiy4,fiz4);\
 910     fix1 = _mm_add_pd(fix1,fiy1);\
 911     fiz1 = _mm_add_pd(fiz1,fix2);\
 912     fiy2 = _mm_add_pd(fiy2,fiz2);\
 913     fix3 = _mm_add_pd(fix3,fiy3);\
 914     fiz3 = _mm_add_pd(fiz3,fix4);\
 915     fiy4 = _mm_add_pd(fiy4,fiz4);\
 916     _mm_storeu_pd( fptr, _mm_add_pd( _mm_loadu_pd(fptr),       fix1 ));\
 917     _mm_storeu_pd( fptr+2, _mm_add_pd( _mm_loadu_pd(fptr+2),   fiz1 ));\
 918     _mm_storeu_pd( fptr+4, _mm_add_pd( _mm_loadu_pd(fptr+4),   fiy2 ));\
 919     _mm_storeu_pd( fptr+6, _mm_add_pd( _mm_loadu_pd(fptr+6),   fix3 ));\
 920     _mm_storeu_pd( fptr+8, _mm_add_pd( _mm_loadu_pd(fptr+8),   fiz3 ));\
 921     _mm_storeu_pd( fptr+10, _mm_add_pd( _mm_loadu_pd(fptr+10), fiy4 ));\
 922     _t1 = _mm_shuffle_pd(fiz1,fiy2,_MM_SHUFFLE2(0,1));\
 923     fix1 = _mm_add_pd(fix1,_t1);\
 924     _t2 = _mm_shuffle_pd(fiz3,fiy4,_MM_SHUFFLE2(0,1));\
 925     fix3 = _mm_add_pd(fix3,_t2);\
 926     fix1 = _mm_add_pd(fix1,fix3);\
 927     fiz1 = _mm_add_sd(fiz1, _mm_unpackhi_pd(fiy2,fiy2));\
 928     fiz3 = _mm_add_sd(fiz3, _mm_unpackhi_pd(fiy4,fiy4));\
 929     fiz1 = _mm_add_sd(fiz1,fiz3);\
 930     _mm_storeu_pd( fshiftptr, _mm_add_pd( _mm_loadu_pd(fshiftptr), fix1 ));\
 931     _mm_store_sd( fshiftptr+2, _mm_add_sd( _mm_load_sd(fshiftptr+2), fiz1 ));\
 932 }
 933 #else
 934 /* Real function for sane compilers */
 935 static gmx_inline void
 936 gmx_mm_update_iforce_4atom_swizzle_pd(__m128d fix1, __m128d fiy1, __m128d fiz1,
 937                                       __m128d fix2, __m128d fiy2, __m128d fiz2,
 938                                       __m128d fix3, __m128d fiy3, __m128d fiz3,
 939                                       __m128d fix4, __m128d fiy4, __m128d fiz4,
 940                                       double * gmx_restrict fptr,
 941                                       double * gmx_restrict fshiftptr)
 942 {
 943     __m128d t1,t2;
 944
 945     /* transpose data */
 946     GMX_MM_TRANSPOSE2_PD(fix1,fiy1);
 947     GMX_MM_TRANSPOSE2_PD(fiz1,fix2);
 948     GMX_MM_TRANSPOSE2_PD(fiy2,fiz2);
 949     GMX_MM_TRANSPOSE2_PD(fix3,fiy3);
 950     GMX_MM_TRANSPOSE2_PD(fiz3,fix4);
 951     GMX_MM_TRANSPOSE2_PD(fiy4,fiz4);
 952
 953     fix1 = _mm_add_pd(fix1,fiy1);
 954     fiz1 = _mm_add_pd(fiz1,fix2);
 955     fiy2 = _mm_add_pd(fiy2,fiz2);
 956     fix3 = _mm_add_pd(fix3,fiy3);
 957     fiz3 = _mm_add_pd(fiz3,fix4);
 958     fiy4 = _mm_add_pd(fiy4,fiz4);
 959
 960     _mm_storeu_pd( fptr, _mm_add_pd( _mm_loadu_pd(fptr),       fix1 ));
 961     _mm_storeu_pd( fptr+2, _mm_add_pd( _mm_loadu_pd(fptr+2),   fiz1 ));
 962     _mm_storeu_pd( fptr+4, _mm_add_pd( _mm_loadu_pd(fptr+4),   fiy2 ));
 963     _mm_storeu_pd( fptr+6, _mm_add_pd( _mm_loadu_pd(fptr+6),   fix3 ));
 964     _mm_storeu_pd( fptr+8, _mm_add_pd( _mm_loadu_pd(fptr+8),   fiz3 ));
 965     _mm_storeu_pd( fptr+10, _mm_add_pd( _mm_loadu_pd(fptr+10), fiy4 ));
 966
 967     t1 = _mm_shuffle_pd(fiz1,fiy2,_MM_SHUFFLE2(0,1));
 968     fix1 = _mm_add_pd(fix1,t1);
 969     t2 = _mm_shuffle_pd(fiz3,fiy4,_MM_SHUFFLE2(0,1));
 970     fix3 = _mm_add_pd(fix3,t2);
 971     fix1 = _mm_add_pd(fix1,fix3); /* x and y sums */
 972
 973     fiz1 = _mm_add_sd(fiz1, _mm_unpackhi_pd(fiy2,fiy2));
 974     fiz3 = _mm_add_sd(fiz3, _mm_unpackhi_pd(fiy4,fiy4));
 975     fiz1 = _mm_add_sd(fiz1,fiz3); /* z sum */
 976
 977     _mm_storeu_pd( fshiftptr, _mm_add_pd( _mm_loadu_pd(fshiftptr), fix1 ));
 978     _mm_store_sd( fshiftptr+2, _mm_add_sd( _mm_load_sd(fshiftptr+2), fiz1 ));
 979 }
 980 #endif
 981
 982
 983 static gmx_inline void
 984 gmx_mm_update_1pot_pd(__m128d pot1, double * gmx_restrict ptrA)
 985 {
 986     pot1 = _mm_add_pd(pot1, _mm_unpackhi_pd(pot1,pot1));
 987     _mm_store_sd(ptrA,_mm_add_sd(pot1,_mm_load_sd(ptrA)));
 988 }
 989
 990 static gmx_inline void
 991 gmx_mm_update_2pot_pd(__m128d pot1, double * gmx_restrict ptrA,
 992                       __m128d pot2, double * gmx_restrict ptrB)
 993 {
 994     GMX_MM_TRANSPOSE2_PD(pot1,pot2);
 995     pot1 = _mm_add_pd(pot1,pot2);
 996     pot2 = _mm_unpackhi_pd(pot1,pot1);
 997
 998     _mm_store_sd(ptrA,_mm_add_sd(pot1,_mm_load_sd(ptrA)));
 999     _mm_store_sd(ptrB,_mm_add_sd(pot2,_mm_load_sd(ptrB)));
1000 }
1001
1002
1003 #endif /* _kernelutil_x86_sse2_double_h_ */