Represents a single lane of a SIMD vector with a runtime-determined index. More...

#include <lane.hpp>

Public Member Functions
ace	Lane (VectorType &vec, const int lane)
	Construct a lane accessor bound to vec at runtime index lane.
ace argon_type	operator= (const scalar_type b)
	Write a scalar value into this lane and return the updated vector.
ace argon_type	Load (const scalar_type *ptr)
	Load a scalar from ptr into this lane and return the updated vector.
ace argon_type	Set (const scalar_type b)
	Set this lane to b and return the updated vector.
ace scalar_type	Get () const
	Get the scalar value of this lane.
ace	operator scalar_type () const
	Read the lane value (implicit conversion to scalar).
ace neon::Vec64_t< scalar_type >	vec ()
	On ARM32, return the 64-bit half-register that contains this lane.
ace int	lane ()
	On ARM32, return the local lane index within the 64-bit half-register returned by vec().

Detailed Description

template<typename VectorType>
class argon::Lane< VectorType >

Represents a single lane of a SIMD vector with a runtime-determined index.

Template Parameters

VectorType The intrinsic SIMD vector type (e.g., int32x4_t).

Provides type-safe read/write access to one element of a SIMD register when the lane index is not known until runtime. On ARM32 (__ARM_ARCH < 8) quadword vectors, vec() returns the 64-bit half-register containing the lane, and lane() returns the local index within that half.

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

include/argon/lane.hpp

Public Member Functions

Detailed Description