gr_fir_fsf_simd Class Reference
[Filters]

common base class for SIMD versions of gr_fir_fsf

This base class handles alignment issues common to SSE and 3DNOW subclasses. More...

#include <gr_fir_fsf_simd.h>

Inheritance diagram for gr_fir_fsf_simd:

[legend]

Collaboration diagram for gr_fir_fsf_simd:

[legend]

List of all members.

Public Member Functions
	gr_fir_fsf_simd ()
	gr_fir_fsf_simd (const std::vector< float > &taps)
	~gr_fir_fsf_simd ()
virtual void	set_taps (const std::vector< float > &taps)
	install new_taps as the current taps.
virtual short	filter (const float input[])
	compute a single output value.
Protected Types
typedef float(*	float_dotprod_t )(const float *input, const float *taps, unsigned n_4_float_blocks)
Protected Attributes
float *	d_aligned_taps [4]
float_dotprod_t	d_float_dotprod

---

Detailed Description

common base class for SIMD versions of gr_fir_fsf

This base class handles alignment issues common to SSE and 3DNOW subclasses.

---

Member Typedef Documentation

typedef float(* gr_fir_fsf_simd::float_dotprod_t)(const float *input, const float *taps, unsigned n_4_float_blocks) [protected]

---

Constructor & Destructor Documentation

gr_fir_fsf_simd::gr_fir_fsf_simd

(

)

gr_fir_fsf_simd::gr_fir_fsf_simd

(

const std::vector< float > &

taps

)

gr_fir_fsf_simd::~gr_fir_fsf_simd

(

)

---

Member Function Documentation

virtual short gr_fir_fsf_simd::filter

(

const float

input[]

)

[virtual]

compute a single output value.

input must have ntaps() valid entries. input[0] .. input[ntaps() - 1] are referenced to compute the output value.

Returns:

the filtered input value.

Reimplemented from gr_fir_fsf_generic.

virtual void gr_fir_fsf_simd::set_taps

(

const std::vector< float > &

taps

)

[virtual]

install new_taps as the current taps.

Reimplemented from gr_fir_fsf.

---

Member Data Documentation

float* gr_fir_fsf_simd::d_aligned_taps[4] [protected]

aligned_taps holds 4 copies of the coefficients preshifted by 0, 1, 2, or 3 floats to meet all possible input data alignments. This allows us to always fetch data and taps that are 128-bit aligned.

float_dotprod_t gr_fir_fsf_simd::d_float_dotprod [protected]

---

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

• gr_fir_fsf_simd.h