2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2009,2010,2012,2013,2014, 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.
43 /*#define GMX_FFT_FFTW3*/
47 #include "gromacs/fft/fft.h"
48 #include "gromacs/math/gmxcomplex.h"
49 #include "gromacs/utility/gmxmpi.h"
55 /*currently only special optimization for FFTE*/
61 #define FFTW(x) fftwf_ ## x
63 #define FFTW(x) fftw_ ## x
71 double fft, local, mpi1, mpi2;
73 typedef struct fft5d_time_t *fft5d_time;
75 #include "gromacs/timing/wallcycle.h"
79 typedef gmx_wallcycle_t fft5d_time;
82 typedef enum fft5d_flags_t {
85 FFT5D_REALCOMPLEX = 4,
94 t_complex *lout, *lout2, *lout3;
95 gmx_fft_t* p1d[3]; /*1D plans*/
97 FFTW(plan) p2d; /*2D plan: used for 1D decomposition if FFT supports transposed output*/
98 FFTW(plan) p3d; /*3D plan: used for 0D decomposition if FFT supports transposed output*/
103 int N[3], M[3], K[3]; /*local length in transposed coordinate system (if not divisisable max)*/
104 int pN[3], pM[3], pK[3]; /*local length - not max but length for this processor*/
105 int oM[3], oK[3]; /*offset for current processor*/
106 int *iNin[3], *oNin[3], *iNout[3], *oNout[3]; /*size for each processor (if divisisable=max) for out(=split)
107 and in (=join) and offsets in transposed coordinate system*/
108 int C[3], rC[3]; /*global length (of the one global axes) */
109 /* C!=rC for real<->complex. then C=rC/2 but with potential padding*/
110 int P[2]; /*size of processor grid*/
113 /* int realcomplex;*/
115 /*int N0,N1,M0,M1,K0,K1;*/
122 typedef struct fft5d_plan_t *fft5d_plan;
124 void fft5d_execute(fft5d_plan plan, int thread, fft5d_time times);
125 fft5d_plan fft5d_plan_3d(int N, int M, int K, MPI_Comm comm[2], int flags, t_complex**lin, t_complex**lin2, t_complex**lout2, t_complex**lout3, int nthreads);
126 void fft5d_local_size(fft5d_plan plan, int* N1, int* M0, int* K0, int* K1, int** coor);
127 void fft5d_destroy(fft5d_plan plan);
128 fft5d_plan fft5d_plan_3d_cart(int N, int M, int K, MPI_Comm comm, int P0, int flags, t_complex** lin, t_complex** lin2, t_complex** lout2, t_complex** lout3, int nthreads);
129 void fft5d_compare_data(const t_complex* lin, const t_complex* in, fft5d_plan plan, int bothLocal, int normarlize);