FFT kernels on real input values

Module: Transform Functions / FFT transforms

Functions

	Name
void	plp_rfft_f32p_xpulpv2(void * arg) Floating-point FFT on real input data for XPULPV2 extension.
void	plp_rfft_f32s_xpulpv2(const plp_fft_instance_f32 * S, const float32_t restrict pSrc, float32_t restrict pDst) Floating-point FFT on real input data for XPULPV2 extension.

Detailed Description

These kernels calculate the FFT transform on real input data. If the dimension of input is FFTLen float32 numbers, the output buffer must contain at least (2*FFTLen) float32 values (used for intermediate computations). Due to the symmetry of real FFT only the first (FFTLen + 2) values of the result are computed in the last stage, corresponding to (FFTLen / 2 + 1) complex values. Supported algorithms: radix-2

These kernels calculate the FFT transform on real input data. If the dimension of input is FFTLen float32 numbers, the output buffer must contain at least (2*FFTLen) float32 values (used for intermediate computations). Due to the symmetry of real FFT only the first (FFTLen + 2) values of the result are computed in the last stage, corresponding to (FFTLen / 2 + 1) complex values. Supported algorithms: radix-2 radix-4 radix-8

Functions Documentation

function plp_rfft_f32p_xpulpv2

void plp_rfft_f32p_xpulpv2(
    void * arg
)

Floating-point FFT on real input data for XPULPV2 extension.

Parameters:

arg points to an instance of the floating-point FFT structure

Return: none

Floating-point FFT on real input data for XPULPV2 extension (parallel version).

function plp_rfft_f32s_xpulpv2

void plp_rfft_f32s_xpulpv2(
    const plp_fft_instance_f32 * S,
    const float32_t *__restrict__ pSrc,
    float32_t *__restrict__ pDst
)

Floating-point FFT on real input data for XPULPV2 extension.

Parameters:

S points to an instance of the floating-point FFT structure
pSrc points to the input buffer (real data)
pDst points to the output buffer (complex data)

Return: none

Updated on 2023-03-01 at 16:16:32 +0000