Chris@10: Next: MPI Data Distribution, Chris@10: Previous: Linking and Initializing MPI FFTW, Chris@10: Up: Distributed-memory FFTW with MPI Chris@10:
Before we document the FFTW MPI interface in detail, we begin with a
Chris@10: simple example outlining how one would perform a two-dimensional
Chris@10: N0 by N1 complex DFT.
Chris@10:
Chris@10:
#include <fftw3-mpi.h>
Chris@10:
Chris@10: int main(int argc, char **argv)
Chris@10: {
Chris@10: const ptrdiff_t N0 = ..., N1 = ...;
Chris@10: fftw_plan plan;
Chris@10: fftw_complex *data;
Chris@10: ptrdiff_t alloc_local, local_n0, local_0_start, i, j;
Chris@10:
Chris@10: MPI_Init(&argc, &argv);
Chris@10: fftw_mpi_init();
Chris@10:
Chris@10: /* get local data size and allocate */
Chris@10: alloc_local = fftw_mpi_local_size_2d(N0, N1, MPI_COMM_WORLD,
Chris@10: &local_n0, &local_0_start);
Chris@10: data = fftw_alloc_complex(alloc_local);
Chris@10:
Chris@10: /* create plan for in-place forward DFT */
Chris@10: plan = fftw_mpi_plan_dft_2d(N0, N1, data, data, MPI_COMM_WORLD,
Chris@10: FFTW_FORWARD, FFTW_ESTIMATE);
Chris@10:
Chris@10: /* initialize data to some function my_function(x,y) */
Chris@10: for (i = 0; i < local_n0; ++i) for (j = 0; j < N1; ++j)
Chris@10: data[i*N1 + j] = my_function(local_0_start + i, j);
Chris@10:
Chris@10: /* compute transforms, in-place, as many times as desired */
Chris@10: fftw_execute(plan);
Chris@10:
Chris@10: fftw_destroy_plan(plan);
Chris@10:
Chris@10: MPI_Finalize();
Chris@10: }
Chris@10:
Chris@10: As can be seen above, the MPI interface follows the same basic style Chris@10: of allocate/plan/execute/destroy as the serial FFTW routines. All of Chris@10: the MPI-specific routines are prefixed with ‘fftw_mpi_’ instead Chris@10: of ‘fftw_’. There are a few important differences, however: Chris@10: Chris@10:
First, we must call fftw_mpi_init() after calling
Chris@10: MPI_Init (required in all MPI programs) and before calling any
Chris@10: other ‘fftw_mpi_’ routine.
Chris@10:
Chris@10:
Chris@10:
Second, when we create the plan with fftw_mpi_plan_dft_2d,
Chris@10: analogous to fftw_plan_dft_2d, we pass an additional argument:
Chris@10: the communicator, indicating which processes will participate in the
Chris@10: transform (here MPI_COMM_WORLD, indicating all processes).
Chris@10: Whenever you create, execute, or destroy a plan for an MPI transform,
Chris@10: you must call the corresponding FFTW routine on all processes
Chris@10: in the communicator for that transform. (That is, these are
Chris@10: collective calls.) Note that the plan for the MPI transform
Chris@10: uses the standard fftw_execute and fftw_destroy routines
Chris@10: (on the other hand, there are MPI-specific new-array execute functions
Chris@10: documented below).
Chris@10:
Chris@10:
Chris@10:
Third, all of the FFTW MPI routines take ptrdiff_t arguments
Chris@10: instead of int as for the serial FFTW. ptrdiff_t is a
Chris@10: standard C integer type which is (at least) 32 bits wide on a 32-bit
Chris@10: machine and 64 bits wide on a 64-bit machine. This is to make it easy
Chris@10: to specify very large parallel transforms on a 64-bit machine. (You
Chris@10: can specify 64-bit transform sizes in the serial FFTW, too, but only
Chris@10: by using the ‘guru64’ planner interface. See 64-bit Guru Interface.)
Chris@10:
Chris@10:
Chris@10:
Fourth, and most importantly, you don't allocate the entire
Chris@10: two-dimensional array on each process. Instead, you call
Chris@10: fftw_mpi_local_size_2d to find out what portion of the
Chris@10: array resides on each processor, and how much space to allocate.
Chris@10: Here, the portion of the array on each process is a local_n0 by
Chris@10: N1 slice of the total array, starting at index
Chris@10: local_0_start. The total number of fftw_complex numbers
Chris@10: to allocate is given by the alloc_local return value, which
Chris@10: may be greater than local_n0 * N1 (in case some
Chris@10: intermediate calculations require additional storage). The data
Chris@10: distribution in FFTW's MPI interface is described in more detail by
Chris@10: the next section.
Chris@10:
Chris@10:
Chris@10:
Given the portion of the array that resides on the local process, it
Chris@10: is straightforward to initialize the data (here to a function
Chris@10: myfunction) and otherwise manipulate it. Of course, at the end
Chris@10: of the program you may want to output the data somehow, but
Chris@10: synchronizing this output is up to you and is beyond the scope of this
Chris@10: manual. (One good way to output a large multi-dimensional distributed
Chris@10: array in MPI to a portable binary file is to use the free HDF5
Chris@10: library; see the HDF home page.)
Chris@10:
Chris@10:
Chris@10:
Chris@10:
Chris@10: