Compute Library
 21.05
NEDirectConvolution3x3.h File Reference
#include <arm_neon.h>

Go to the source code of this file.

Namespaces

 arm_compute
 Copyright (c) 2017-2021 Arm Limited.
 
 arm_compute::detail
 

Functions

float32x4x3_t load_matrix_row (const float *ptr)
 
template<unsigned int stridex>
float32x4x2_t convolve_3x3 (const float *in_top, const float *in_mid, const float *in_low, const float32x4x3_t &m0, const float32x4x3_t &m1, const float32x4x3_t &m2)
 
template<>
float32x4x2_t convolve_3x3< 1 > (const float *in_top, const float *in_mid, const float *in_low, const float32x4x3_t &m0, const float32x4x3_t &m1, const float32x4x3_t &m2)
 
template<>
float32x4x2_t convolve_3x3< 2 > (const float *in_top, const float *in_mid, const float *in_low, const float32x4x3_t &m0, const float32x4x3_t &m1, const float32x4x3_t &m2)
 
template<>
float32x4x2_t convolve_3x3< 3 > (const float *in_top, const float *in_mid, const float *in_low, const float32x4x3_t &m0, const float32x4x3_t &m1, const float32x4x3_t &m2)
 
template<unsigned int stridex>
void store_results (float *buffer, const float32x4x2_t &values)
 Stores a float32x4x2_t array into a memory location. More...
 
template<>
void store_results< 1 > (float *buffer, const float32x4x2_t &values)
 
template<>
void store_results< 2 > (float *buffer, const float32x4x2_t &values)
 
template<>
void store_results< 3 > (float *buffer, const float32x4x2_t &values)
 
template<unsigned int stridex>
int get_input_num_elems_processed (unsigned int num_elems_written_per_iteration)
 
template<>
int get_input_num_elems_processed< 1 > (unsigned int num_elems_written_per_iteration)
 
template<>
int get_input_num_elems_processed< 2 > (unsigned int num_elems_written_per_iteration)
 
template<>
int get_input_num_elems_processed< 3 > (unsigned int num_elems_written_per_iteration)