Real Matrix

Data Structures
struct	msp_matrix_add_q15_params
	Parameter structure for the matrix add function. More...
struct	msp_matrix_add_iq31_params
	Parameter structure for the matrix add function. More...
struct	msp_matrix_sub_q15_params
	Parameter structure for the matrix subtract function. More...
struct	msp_matrix_sub_iq31_params
	Parameter structure for the matrix subtract function. More...
struct	msp_matrix_trans_q15_params
	Parameter structure for the matrix transpose function. More...
struct	msp_matrix_trans_iq31_params
	Parameter structure for the matrix transpose function. More...
struct	msp_matrix_mpy_q15_params
	Parameter structure for the real matrix multiply function. More...
struct	msp_matrix_mpy_iq31_params
	Parameter structure for the real matrix multiply function. More...
struct	msp_matrix_neg_q15_params
	Parameter structure for the real matrix negate function. More...
struct	msp_matrix_neg_iq31_params
	Parameter structure for the real matrix negate function. More...
struct	msp_matrix_abs_q15_params
	Parameter structure for the real matrix absolute value function. More...
struct	msp_matrix_abs_iq31_params
	Parameter structure for the real matrix absolute value function. More...
struct	msp_matrix_offset_q15_params
	Parameter structure for the real matrix offset function. More...
struct	msp_matrix_offset_iq31_params
	Parameter structure for the real matrix offset function. More...
struct	msp_matrix_scale_q15_params
	Parameter structure for the real matrix scale function. More...
struct	msp_matrix_scale_iq31_params
	Parameter structure for the real matrix scale function. More...
struct	msp_matrix_shift_q15_params
	Parameter structure for the real matrix shift function. More...
struct	msp_matrix_shift_iq31_params
	Parameter structure for the real matrix shift function. More...

Functions
msp_status	msp_matrix_add_q15 (const msp_matrix_add_q15_params *params, const _q15 *srcA, const _q15 *srcB, _q15 *dst)
	Addition of two real source matrices. More...
msp_status	msp_matrix_add_iq31 (const msp_matrix_add_iq31_params *params, const _iq31 *srcA, const _iq31 *srcB, _iq31 *dst)
	Addition of two real source matrices. More...
msp_status	msp_matrix_sub_q15 (const msp_matrix_sub_q15_params *params, const _q15 *srcA, const _q15 *srcB, _q15 *dst)
	Subtraction of two real source matrices. More...
msp_status	msp_matrix_sub_iq31 (const msp_matrix_sub_iq31_params *params, const _iq31 *srcA, const _iq31 *srcB, _iq31 *dst)
	Subtraction of two real source matrices. More...
msp_status	msp_matrix_mpy_q15 (const msp_matrix_mpy_q15_params *params, const _q15 *srcA, const _q15 *srcB, _q15 *dst)
	Multiplication of two real source matrices. More...
msp_status	msp_matrix_mpy_iq31 (const msp_matrix_mpy_iq31_params *params, const _iq31 *srcA, const _iq31 *srcB, _iq31 *dst)
	Multiplication of two real source matrices. More...
msp_status	msp_matrix_trans_q15 (const msp_matrix_trans_q15_params *params, const _q15 *src, _q15 *dst)
	Transposition of a source matrix. More...
msp_status	msp_matrix_trans_iq31 (const msp_matrix_trans_iq31_params *params, const _iq31 *src, _iq31 *dst)
	Transposition of a source matrix. More...
msp_status	msp_matrix_neg_q15 (const msp_matrix_neg_q15_params *params, const _q15 *src, _q15 *dst)
	Negation of a source matrix. More...
msp_status	msp_matrix_neg_iq31 (const msp_matrix_neg_iq31_params *params, const _iq31 *src, _iq31 *dst)
	Negation of a source matrix. More...
msp_status	msp_matrix_abs_q15 (const msp_matrix_abs_q15_params *params, const _q15 *src, _q15 *dst)
	Absolute value of a real source matrix. More...
msp_status	msp_matrix_abs_iq31 (const msp_matrix_abs_iq31_params *params, const _iq31 *src, _iq31 *dst)
	Absolute value of a real source matrix. More...
msp_status	msp_matrix_offset_q15 (const msp_matrix_offset_q15_params *params, const _q15 *src, _q15 *dst)
	Constant offset of a real source matrix. More...
msp_status	msp_matrix_offset_iq31 (const msp_matrix_offset_iq31_params *params, const _iq31 *src, _iq31 *dst)
	Constant offset of a real source matrix. More...
msp_status	msp_matrix_scale_q15 (const msp_matrix_scale_q15_params *params, const _q15 *src, _q15 *dst)
	Scale a real source matrix. More...
msp_status	msp_matrix_scale_iq31 (const msp_matrix_scale_iq31_params *params, const _iq31 *src, _iq31 *dst)
	Scale a real source matrix. More...
msp_status	msp_matrix_shift_q15 (const msp_matrix_shift_q15_params *params, const _q15 *src, _q15 *dst)
	Bitwise shift of a real source matrix. More...
msp_status	msp_matrix_shift_iq31 (const msp_matrix_shift_iq31_params *params, const _iq31 *src, _iq31 *dst)
	Bitwise shift of a real source matrix. More...

Detailed Description

Functions for performing matrix operations on real data.

Function Documentation

msp_status msp_matrix_abs_iq31	(	const msp_matrix_abs_iq31_params *	params,
		const _iq31 *	src,
		_iq31 *	dst
	)

Absolute value of a real source matrix.

Details

Element-wise absolute value of a single source matrix.

Pseudo code

dst = |src|

LEA Support

This function is not currently supported by LEA.

Parameters

params	Pointer to the matrix absolute value structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_abs_q15	(	const msp_matrix_abs_q15_params *	params,
		const _q15 *	src,
		_q15 *	dst
	)

Absolute value of a real source matrix.

Details

Element-wise absolute value of a single source matrix.

Pseudo code

dst = |src|

LEA Support

This function is not currently supported by LEA.

Parameters

params	Pointer to the matrix absolute value structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_add_iq31	(	const msp_matrix_add_iq31_params *	params,
		const _iq31 *	srcA,
		const _iq31 *	srcB,
		_iq31 *	dst
	)

Addition of two real source matrices.

Details

Element-wise IQ31 addition of two real matrices with saturation. This function supports in-place operations.

Pseudo code

dst = srcA + srcB

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix add parameter structure.
srcA	Pointer to the first source data matrix.
srcB	Pointer to the second source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_add_q15	(	const msp_matrix_add_q15_params *	params,
		const _q15 *	srcA,
		const _q15 *	srcB,
		_q15 *	dst
	)

Addition of two real source matrices.

Details

Element-wise Q15 addition of two real matrices with saturation. This function supports in-place operations.

Pseudo code

dst = srcA + srcB

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix add parameter structure.
srcA	Pointer to the first source data matrix.
srcB	Pointer to the second source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

Examples:

matrix_ex1_add_q15.c.

msp_status msp_matrix_mpy_iq31	(	const msp_matrix_mpy_iq31_params *	params,
		const _iq31 *	srcA,
		const _iq31 *	srcB,
		_iq31 *	dst
	)

Multiplication of two real source matrices.

Details

Element-wise IQ31 multiplication of two real source matrices without saturation. This function supports in-place operation.

Pseudo code

dst = srcA * srcB

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix multiply parameter structure.
srcA	Pointer to the first source data matrix.
srcB	Pointer to the second source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_mpy_q15	(	const msp_matrix_mpy_q15_params *	params,
		const _q15 *	srcA,
		const _q15 *	srcB,
		_q15 *	dst
	)

Multiplication of two real source matrices.

Details

Element-wise Q15 multiplication of two real source matrices without saturation. This function supports in-place operation.

Pseudo code

dst = srcA * srcB

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix multiply parameter structure.
srcA	Pointer to the first source data matrix.
srcB	Pointer to the second source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

Examples:

matrix_ex3_mpy_q15.c.

msp_status msp_matrix_neg_iq31	(	const msp_matrix_neg_iq31_params *	params,
		const _iq31 *	src,
		_iq31 *	dst
	)

Negation of a source matrix.

Details

Element-wise IQ31 multiplication by negative one with a real source matrix. This function supports in-place operation.

Pseudo code

dst = -src

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix negate parameter structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_neg_q15	(	const msp_matrix_neg_q15_params *	params,
		const _q15 *	src,
		_q15 *	dst
	)

Negation of a source matrix.

Details

Element-wise Q15 multiplication by negative one with a real source matrix. This function supports in-place operation.

Pseudo code

dst = -src

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix negate parameter structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_offset_iq31	(	const msp_matrix_offset_iq31_params *	params,
		const _iq31 *	src,
		_iq31 *	dst
	)

Constant offset of a real source matrix.

Details

Element-wise IQ31 addition with saturation of a single source matrix with a IQ31 constant. This function supports in-place operation.

Pseudo code

dst = src + offset

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix offset structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_offset_q15	(	const msp_matrix_offset_q15_params *	params,
		const _q15 *	src,
		_q15 *	dst
	)

Constant offset of a real source matrix.

Details

Element-wise Q15 addition with saturation of a single source matrix with a Q15 constant. This function supports in-place operation.

Pseudo code

dst = src + offset

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix offset structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_scale_iq31	(	const msp_matrix_scale_iq31_params *	params,
		const _iq31 *	src,
		_iq31 *	dst
	)

Scale a real source matrix.

Details

Element-wise scaling of a real matrix. Source data is multiplied by a IQ31 constant and then shifted left. This function supports in-place operation.

Pseudo code

dst = src * 2^SHIFT * SCALE

LEA Support

This function is not currently supported by LEA.

Parameters

params	Pointer to the matrix scale structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_scale_q15	(	const msp_matrix_scale_q15_params *	params,
		const _q15 *	src,
		_q15 *	dst
	)

Scale a real source matrix.

Details

Element-wise scaling of a real matrix. Source data is multiplied by a IQ31 constant and then shifted left. This function supports in-place operation.

Pseudo code

dst = src * 2^SHIFT * SCALE

LEA Support

This function is not currently supported by LEA.

Parameters

params	Pointer to the matrix scale structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_shift_iq31	(	const msp_matrix_shift_iq31_params *	params,
		const _iq31 *	src,
		_iq31 *	dst
	)

Bitwise shift of a real source matrix.

Details

Element-wise bitwise shift of a real matrix to the left or right by a signed integer value. This function does not saturate and supports in-place operation.

Pseudo code

dst = src * 2^SHIFT

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix shift structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_shift_q15	(	const msp_matrix_shift_q15_params *	params,
		const _q15 *	src,
		_q15 *	dst
	)

Bitwise shift of a real source matrix.

Details

Element-wise bitwise shift of a real matrix to the left or right by a signed integer value. This function does not saturate and supports in-place operation.

Pseudo code

dst = src * 2^SHIFT

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix shift structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_sub_iq31	(	const msp_matrix_sub_iq31_params *	params,
		const _iq31 *	srcA,
		const _iq31 *	srcB,
		_iq31 *	dst
	)

Subtraction of two real source matrices.

Details

Element-wise IQ31 subtraction of two real matrices with saturation. This function supports in-place operations.

Pseudo code

dst = srcA - srcB

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix subtraction parameter structure.
srcA	Pointer to the first source data matrix.
srcB	Pointer to the second source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_sub_q15	(	const msp_matrix_sub_q15_params *	params,
		const _q15 *	srcA,
		const _q15 *	srcB,
		_q15 *	dst
	)

Subtraction of two real source matrices.

Details

Element-wise Q15 subtraction of two real matrices with saturation. This function supports in-place operations.

Pseudo code

dst = srcA - srcB

LEA Support

This function is supported by LEA and requires data to be placed into shared LEA memory with 4-byte alignment.

Parameters

params	Pointer to the matrix subtraction parameter structure.
srcA	Pointer to the first source data matrix.
srcB	Pointer to the second source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_trans_iq31	(	const msp_matrix_trans_iq31_params *	params,
		const _iq31 *	src,
		_iq31 *	dst
	)

Transposition of a source matrix.

Details

Transposition of the source matrix. The resulting matrix will have same number of rows as the source does columns and the same number of columns as the source has rows. This function does not support in-place operations.

Pseudo code

dst = src'

LEA Support

This function is not currently supported by LEA.

Parameters

params	Pointer to the matrix transpose parameter structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

msp_status msp_matrix_trans_q15	(	const msp_matrix_trans_q15_params *	params,
		const _q15 *	src,
		_q15 *	dst
	)

Transposition of a source matrix.

Details

Transposition of the source matrix. The resulting matrix will have same number of rows as the source does columns and the same number of columns as the source has rows. This function does not support in-place operations.

Pseudo code

dst = src'

LEA Support

This function is not currently supported by LEA.

Parameters

params	Pointer to the matrix transpose parameter structure.
src	Pointer to the source data matrix.
dst	Pointer to the destination data matrix.

Returns

Status of the operation.

Examples:

matrix_ex2_trans_q15.c.