FK20 CUDA
fk20_poly2h_fft.cu File Reference
#include <time.h>
#include <stdio.h>
#include "fr.cuh"
#include "g1.cuh"
#include "fk20.cuh"
Include dependency graph for fk20_poly2h_fft.cu:

Go to the source code of this file.

Macros

#define ROWS   512
 

Functions

__global__ void fk20_hext2h (g1p_t *h)
 hext -> h Fill upper half of hext with inf, modifying in place. More...
 
__host__ void fk20_poly2h_fft (g1p_t *h_fft, const fr_t *polynomial, const g1p_t xext_fft[8192], unsigned rows)
 polynomial + xext_fft -> h_fft This function is a wrapper for the full FK20 computation, i.e. commits up to 512 polynomials of 4096 elements to the same setup. l = 16, intrinsic to the implementation. More...
 

Macro Definition Documentation

◆ ROWS

#define ROWS   512

Definition at line 13 of file fk20_poly2h_fft.cu.

Function Documentation

◆ fk20_hext2h()

__global__ void fk20_hext2h ( g1p_t h)

hext -> h Fill upper half of hext with inf, modifying in place.

Parameters
[in,out]h
Returns
void

Definition at line 27 of file fk20_poly2h_fft.cu.

Here is the call graph for this function:

◆ fk20_poly2h_fft()

__host__ void fk20_poly2h_fft ( g1p_t h_fft,
const fr_t polynomial,
const g1p_t  xext_fft[8192],
unsigned  rows 
)

polynomial + xext_fft -> h_fft This function is a wrapper for the full FK20 computation, i.e. commits up to 512 polynomials of 4096 elements to the same setup. l = 16, intrinsic to the implementation.

Parameters
[out]h_fftarray with dimensions [rows * 512]
[in]polynomialarray with dimensions [rows * 16 * 512]
[in]xext_fftarray with dimensions [16 * 512]
[in]rowsnumber of rows (gridDim.x)
Returns
void

Definition at line 47 of file fk20_poly2h_fft.cu.

Here is the call graph for this function:
Here is the caller graph for this function: