2 * This file is part of the GROMACS molecular simulation package.
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.
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.
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.
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.
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.
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.
35 #include "gromacs/pbcutil/ishift.h"
36 #include "gromacs/simd/simd.h"
37 #include "gromacs/simd/simd_math.h"
38 #include "gromacs/simd/vector_operations.h"
39 #include "../../nbnxn_consts.h"
40 #ifdef CALC_COUL_EWALD
41 #include "gromacs/math/utilities.h"
44 #ifndef GMX_SIMD_J_UNROLL_SIZE
45 #error "Need to define GMX_SIMD_J_UNROLL_SIZE before including the 4xn kernel common header file"
48 #define UNROLLI NBNXN_CPU_CLUSTER_I_SIZE
49 #define UNROLLJ (GMX_SIMD_REAL_WIDTH/GMX_SIMD_J_UNROLL_SIZE)
51 /* The stride of all the atom data arrays is max(UNROLLI,unrollj) */
52 #if GMX_SIMD_REAL_WIDTH >= UNROLLI
53 #define STRIDE (GMX_SIMD_REAL_WIDTH/GMX_SIMD_J_UNROLL_SIZE)
55 #define STRIDE (UNROLLI)
58 #include "../nbnxn_kernel_simd_utils.h"
60 static gmx_inline void
61 gmx_load_simd_4xn_interactions(int gmx_unused excl,
62 gmx_exclfilter gmx_unused filter_S0,
63 gmx_exclfilter gmx_unused filter_S1,
64 gmx_exclfilter gmx_unused filter_S2,
65 gmx_exclfilter gmx_unused filter_S3,
66 const char gmx_unused *interaction_mask_indices,
67 real gmx_unused *simd_interaction_array,
68 gmx_simd_bool_t *interact_S0,
69 gmx_simd_bool_t *interact_S1,
70 gmx_simd_bool_t *interact_S2,
71 gmx_simd_bool_t *interact_S3)
73 #if defined GMX_SIMD_X86_SSE2_OR_HIGHER || defined GMX_SIMD_REFERENCE
74 /* Load integer interaction mask */
75 gmx_exclfilter mask_pr_S = gmx_load1_exclfilter(excl);
76 *interact_S0 = gmx_checkbitmask_pb(mask_pr_S, filter_S0);
77 *interact_S1 = gmx_checkbitmask_pb(mask_pr_S, filter_S1);
78 *interact_S2 = gmx_checkbitmask_pb(mask_pr_S, filter_S2);
79 *interact_S3 = gmx_checkbitmask_pb(mask_pr_S, filter_S3);
81 #ifdef GMX_SIMD_IBM_QPX
82 const int size = GMX_SIMD_REAL_WIDTH * sizeof(real);
83 *interact_S0 = gmx_load_interaction_mask_pb(size*interaction_mask_indices[0], simd_interaction_array);
84 *interact_S1 = gmx_load_interaction_mask_pb(size*interaction_mask_indices[1], simd_interaction_array);
85 *interact_S2 = gmx_load_interaction_mask_pb(size*interaction_mask_indices[2], simd_interaction_array);
86 *interact_S3 = gmx_load_interaction_mask_pb(size*interaction_mask_indices[3], simd_interaction_array);
90 /* All functionality defines are set here, except for:
91 * CALC_ENERGIES, ENERGY_GROUPS which are defined before.
92 * CHECK_EXCLS, which is set just before including the inner loop contents.
93 * The combination rule defines, LJ_COMB_GEOM or LJ_COMB_LB are currently
94 * set before calling the kernel function. We might want to move that
95 * to inside the n-loop and have a different combination rule for different
96 * ci's, as no combination rule gives a 50% performance hit for LJ.
99 /* We always calculate shift forces, because it's cheap anyhow */
100 #define CALC_SHIFTFORCES
102 /* Assumes all LJ parameters are identical */
103 /* #define FIX_LJ_C */