OneAPI already has the fft algorithm in mkl, and the main purpose of adding the Intel compiler CI is to ensure that the code conforms to the Fortran standards.
In fact, when testing fftpack for usability with Intel compilers, I found some old programming paradigms:
Passing floating-point type (Wsave(2*n+1:)) to integer (Ifac(*)) might trigger a compiler interface check error.
Description
OneAPI already has the fft algorithm in mkl, and the main purpose of adding the Intel compiler CI is to ensure that the code conforms to the Fortran standards.
In fact, when testing
fftpack
for usability with Intel compilers, I found some old programming paradigms:Passing floating-point type (
Wsave(2*n+1:)
) to integer (Ifac(*)
) might trigger a compiler interface check error.