Square Root

Functions
csky_status	csky_sqrt_f32 (float32_t in, float32_t *pOut)
	Floating-point square root function. More...
csky_status	csky_sqrt_q15 (q15_t in, q15_t *pOut)
	Q15 square root function. More...
csky_status	csky_sqrt_q31 (q31_t in, q31_t *pOut)
	Q31 square root function. More...

Description

Computes the square root of a number. There are separate functions for Q15, Q31, and floating-point data types. When a CPU with FPU is used, the instruction fsqrts is used to compute the result, while a CPU without FPU, Newton-Raphson algorithm is used. For Newton-Raphson algorithm, this is an iterative algorithm of the form:

     x1 = x0 - f(x0)/f'(x0)

where x1 is the current estimate, x0 is the previous estimate, and f'(x0) is the derivative of f() evaluated at x0. For the square root function, the algorithm reduces to:

    x0 = in/2                         [initial guess]
    x1 = 1/2 * ( x0 + in / x0)        [each iteration]

Function Documentation

csky_status csky_sqrt_f32	(	float32_t	in,
		float32_t *	pOut
	)

Parameters

[in]	in	input value.
[out]	pOut	square root of input value.

Returns

The function returns CSKY_MATH_SUCCESS if input value is positive value or CSKY_MATH_ARGUMENT_ERROR if in is negative value and returns zero output for negative values.

csky_status csky_sqrt_q15	(	q15_t	in,
		q15_t *	pOut
	)

Parameters

[in]	in	input value. The range of the input value is [0 +1) or 0x0000 to 0x7FFF.
[out]	*pOut	square root of input value.

Returns

The function returns CSKY_MATH_SUCCESS if the input value is positive and CSKY_MATH_ARGUMENT_ERROR if the input is negative. For negative inputs, the function returns *pOut = 0.

csky_status csky_sqrt_q31	(	q31_t	in,
		q31_t *	pOut
	)

Parameters

[in]	in	input value. The range of the input value is [0 +1) or 0x00000000 to 0x7FFFFFFF.
[out]	*pOut	square root of input value.

Returns

The function returns CSKY_MATH_SUCCESS if the input value is positive and CSKY_MATH_ARGUMENT_ERROR if the input is negative. For negative inputs, the function returns *pOut = 0.

Note

When the hard float instruction fsqrts is used, the accuracy will be lost from 3 LSB to 7 LSB. So as the functions, who call this function.