core_matrix.c

Summary

core_matrix.c
Description	Matrix manipulation benchmark
Functions
core_bench_matrix	Benchmark function
matrix_test	Perform matrix manipulation.
matrix_sum	Calculate a function that depends on the values of elements in the matrix.
matrix_mul_const	Multiply a matrix by a constant.
matrix_add_const	Add a constant value to all elements of a matrix.
matrix_mul_vect	Multiply a matrix by a vector.
matrix_mul_matrix	Multiply a matrix by a matrix.
matrix_mul_matrix_bitextract	Multiply a matrix by a matrix, and extract some bits from the result.

Description

Matrix manipulation benchmark

This very simple algorithm forms the basis of many more complex algorithms.

The tight inner loop is the focus of many optimizations (compiler as well as hardware based) and is thus relevant for embedded processing.

The total available data space will be divided to 3 parts

NxN Matrix A	initialized with small values (upper 3/4 of the bits all zero).
NxN Matrix B	initialized with medium values (upper half of the bits all zero).
NxN Matrix C	used for the result.

The actual values for A and B must be derived based on input that is not available at compile time.

Functions

core_bench_matrix

ee_u16 core_bench_matrix( mat_params * p,
ee_s16 seed,
ee_u16 crc )

Benchmark function

Iterate matrix_test N times, changing the matrix values slightly by a constant amount each time.

matrix_test

ee_s16 matrix_test( ee_u32 N,
MATRES * C,
MATDAT * A,
MATDAT * B,
MATDAT val )

Perform matrix manipulation.

Parameters

N	Dimensions of the matrix.
C	memory for result matrix.
A	input matrix
B	operator matrix (not changed during operations)

Returns

A CRC value that captures all results calculated in the function. In particular, crc of the value calculated on the result matrix after each step by matrix_sum.

Operation

1	Add a constant value to all elements of a matrix.
2	Multiply a matrix by a constant.
3	Multiply a matrix by a vector.
4	Multiply a matrix by a matrix.
5	Add a constant value to all elements of a matrix.

After the last step, matrix A is back to original contents.

matrix_sum

ee_s16 matrix_sum( ee_u32 N,
MATRES * C,
MATDAT clipval )

Calculate a function that depends on the values of elements in the matrix.

For each element, accumulate into a temporary variable.

As long as this value is under the parameter clipval, add 1 to the result if the element is bigger then the previous.

Otherwise, reset the accumulator and add 10 to the result.

matrix_mul_const

void matrix_mul_const( ee_u32 N,
MATRES * C,
MATDAT * A,
MATDAT val )

Multiply a matrix by a constant. This could be used as a scaler for instance.

matrix_add_const

void matrix_add_const( ee_u32 N,
MATDAT * A,
MATDAT val )

Add a constant value to all elements of a matrix.

matrix_mul_vect

void matrix_mul_vect( ee_u32 N,
MATRES * C,
MATDAT * A,
MATDAT * B )

Multiply a matrix by a vector. This is common in many simple filters (e.g. fir where a vector of coefficients is applied to the matrix.)

matrix_mul_matrix

void matrix_mul_matrix( ee_u32 N,
MATRES * C,
MATDAT * A,
MATDAT * B )

Multiply a matrix by a matrix. Basic code is used in many algorithms, mostly with minor changes such as scaling.

matrix_mul_matrix_bitextract

void matrix_mul_matrix_bitextract( ee_u32 N,
MATRES * C,
MATDAT * A,
MATDAT * B )

Multiply a matrix by a matrix, and extract some bits from the result. Basic code is used in many algorithms, mostly with minor changes such as scaling.