d@0: d@0: d@0: Installation and Supported Hardware/Software - FFTW 3.2.1 d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0:
d@0:

d@0: d@0: d@0: Next: , d@0: Previous: Multi-threaded FFTW, d@0: Up: Multi-threaded FFTW d@0:


d@0:
d@0: d@0:

5.1 Installation and Supported Hardware/Software

d@0: d@0:

All of the FFTW threads code is located in the threads d@0: subdirectory of the FFTW package. On Unix systems, the FFTW threads d@0: libraries and header files can be automatically configured, compiled, d@0: and installed along with the uniprocessor FFTW libraries simply by d@0: including --enable-threads in the flags to the configure d@0: script (see Installation on Unix). d@0: d@0: The threads routines require your operating system to have some sort d@0: of shared-memory threads support. Specifically, the FFTW threads d@0: package works with POSIX threads (available on most Unix variants, d@0: from GNU/Linux to MacOS X) and Win32 threads. We also support using d@0: OpenMP, enabled by using d@0: --enable-openmp (instead of --enable-threads). d@0: (This may be useful if you are employing that sort of directive in d@0: your own code, in order to minimize conflicts.) If you have a d@0: shared-memory machine that uses a different threads API, it should be d@0: a simple matter of programming to include support for it; see the file d@0: threads/threads.c for more detail. d@0: d@0:

Ideally, of course, you should also have multiple processors in order to d@0: get any benefit from the threaded transforms. d@0: d@0: d@0: d@0: