GEMMConfigReshaped Struct Reference

GEMM Configuration for Reshaped kernel. More...

#include <Common.h>

Data Fields
unsigned int	m0 { 1 }
	Number of rows processed by the matrix multiplication. More...
unsigned int	n0 { 1 }
	Number of columns processed by the matrix multiplication. More...
unsigned int	k0 { 1 }
	Number of partial accumulations performed by the matrix multiplication. More...
unsigned int	v0 { 1 }
	Number of vertical blocks of size (m0xk0) stored on the same output row. More...
unsigned int	h0 { 1 }
	Number of horizontal blocks of size (k0xn0) stored on the same output row. More...
bool	interleave_lhs { false }
	True if the v0 (m0xk0) blocks have to be interleaved in the output row. More...
bool	interleave_rhs { false }
	True if the h0 (k0xn0) blocks have to be interleaved in the output row. More...
bool	transpose_rhs { false }
	True if the (k0xn0) block has to be transposed before been stored. More...
bool	export_cl_image { false }
	True if the reshaped rhs has to be exported to cl_image. More...

Detailed Description

GEMM Configuration for Reshaped kernel.

Definition at line 66 of file Common.h.

Field Documentation

export_cl_image

bool export_cl_image { false }

True if the reshaped rhs has to be exported to cl_image.

n0 must be equal to 4

Definition at line 76 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

h0

unsigned int h0 { 1 }

Number of horizontal blocks of size (k0xn0) stored on the same output row.

Definition at line 72 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

interleave_lhs

bool interleave_lhs { false }

True if the v0 (m0xk0) blocks have to be interleaved in the output row.

Definition at line 73 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

interleave_rhs

bool interleave_rhs { false }

True if the h0 (k0xn0) blocks have to be interleaved in the output row.

Definition at line 74 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

k0

unsigned int k0 { 1 }

Number of partial accumulations performed by the matrix multiplication.

Definition at line 70 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

m0

unsigned int m0 { 1 }

Number of rows processed by the matrix multiplication.

Definition at line 68 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

n0

unsigned int n0 { 1 }

Number of columns processed by the matrix multiplication.

Definition at line 69 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

transpose_rhs

bool transpose_rhs { false }

True if the (k0xn0) block has to be transposed before been stored.

Definition at line 75 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

v0

unsigned int v0 { 1 }

Number of vertical blocks of size (m0xk0) stored on the same output row.

Definition at line 71 of file Common.h.

Referenced by arm_compute::mlgo::operator<<(), and arm_compute::mlgo::operator==().

---

The documentation for this struct was generated from the following file:

• src/runtime/CL/mlgo/Common.h