Chris@19: Previous: Precision, Chris@19: Up: Data Types and Files Chris@19:
void *fftw_malloc(size_t n); Chris@19: void fftw_free(void *p); Chris@19:Chris@19:
Chris@19: These are functions that behave identically to malloc and
Chris@19: free, except that they guarantee that the returned pointer obeys
Chris@19: any special alignment restrictions imposed by any algorithm in FFTW
Chris@19: (e.g. for SIMD acceleration). See SIMD alignment and fftw_malloc.
Chris@19:
Chris@19:
Chris@19:
Data allocated by fftw_malloc must be deallocated by
Chris@19: fftw_free and not by the ordinary free.
Chris@19:
Chris@19:
These routines simply call through to your operating system's
Chris@19: malloc or, if necessary, its aligned equivalent
Chris@19: (e.g. memalign), so you normally need not worry about any
Chris@19: significant time or space overhead. You are not required to use
Chris@19: them to allocate your data, but we strongly recommend it.
Chris@19:
Chris@19:
Note: in C++, just as with ordinary malloc, you must typecast
Chris@19: the output of fftw_malloc to whatever pointer type you are
Chris@19: allocating.
Chris@19:
Chris@19:
Chris@19:
We also provide the following two convenience functions to allocate
Chris@19: real and complex arrays with n elements, which are equivalent
Chris@19: to (double *) fftw_malloc(sizeof(double) * n) and
Chris@19: (fftw_complex *) fftw_malloc(sizeof(fftw_complex) * n),
Chris@19: respectively:
Chris@19:
Chris@19:
double *fftw_alloc_real(size_t n); Chris@19: fftw_complex *fftw_alloc_complex(size_t n); Chris@19:Chris@19:
Chris@19: The equivalent functions in other precisions allocate arrays of n
Chris@19: elements in that precision. e.g. fftwf_alloc_real(n) is
Chris@19: equivalent to (float *) fftwf_malloc(sizeof(float) * n).
Chris@19:
Chris@19:
Chris@19:
Chris@19:
Chris@19: