[Numpy-discussion] Enhancements for NumPy's FFTs
Sat Mar 14 23:44:04 CDT 2009
> I think MSVC uses _inline
No, MSVC uses a double underscore. That is, __restrict for variable names
and __declspec(restrict) for function return values.
>> #if (__STDC_VERSION__ >= 199901L)
>> #undef RESTRICT
>> #undef INLINE
>> #define RESTRICT restrict
>> #define INLINE inline
> What does this last bit do? We implicitly assume that ieee floating point
It uses the restrict keyword in C99. The last test will pass on any C99
>> #ifdef DOUBLE
>> typedef double Treal
>> typedef float Treal
>> typedef Treal *RESTRICT Vreal /* V as in "vector" */
> I'm not sure about the names. I would prefer to keep the declarations
> the lines of
> double * NPY_RESTRICT ptmp;
Yes, but then I would have to go in and edit all function bodies in
fftpack.c as well. fftpack.c currently uses "Treal array" as naming
convention, so I just changed that to "Vreal array".
I changed all ints to long for 64 bit support.
Well, it compiles and runs ok on my computer now. I'll open a ticket for
the FFT. I'll attach the C files to the ticket.
More information about the Numpy-discussion