FFTW FAQ - Section 5
Known bugs
There was a bug in rfftwnd
causing an incorrect amount of memory to be allocated. The bug showed up in Linux with libc-5.3.12
(and nowhere else that we know of), and has been fixed in FFTW
1.2.
The MPI transforms (really, just the transpose routines) in FFTW 1.2
had bugs that could cause errors in some situations. These bugs have
been corrected in FFTW 1.2.1.
This bug was fixed* in FFTW 1.3. (Older versions of FFTW did
work in single precision, but the test programs didn't--the error
tolerances in the tests were set for double precision.)
* The MPI distributed-memory transforms must be modified to work in
single precision. In the transpose_mpi
directory, replace all occurrences of double
with float
(or whatever precision you are using), and replace
MPI_DOUBLE
with MPI_FLOAT
(or the appropriate type constant for your
precision).
This bug was fixed in FFTW 1.3. FFTW 1.2.1 produced the right answer,
but the test program was wrong. For large n, n*n in the naive
transform that we used for comparison overflows 32 bit integer
precision, breaking the test.
We had problems with glibc-2.0.5. The code should work with
glibc-2.0.7.
This bug was fixed in FFTW 2.0.1. (There was a 32-bit integer
overflow due to a poorly-parenthesized expression.)
Back: Internals of FFTW.
Return to contents.
Matteo Frigo and Steven G. Johnson / fftw@theory.lcs.mit.edu
- 28 September 1998
Extracted from FFTW Frequently Asked Questions with Answers,
Copyright © 1998 Massachusetts Institute of Technology.