SIMD Specific Operations

Detailed Description

Proper SIMD operations as shuffling, splitting and merging SIMD vectors

Variables
constexpr auto	eve::broadcast = functor<broadcast_t>
	Computes the.
constexpr auto	eve::combine = functor<combine_t>
	Combines two SIMD values in a larger one.
constexpr auto	eve::gather = functor<gather_t>
	Load a SIMD value with values selected from a memory region at the given offsets.
constexpr auto	eve::has_equal_in = functor<has_equal_in_t>
	Given two simd_values: x, match_against returns a logical mask. The res[i] == eve::any(x[i] == match_against);.
constexpr auto	eve::iterate_selected = functor<iterate_selected_t>
	a utility to do scalar iteration over all true elements in a logical.
constexpr auto	eve::scan = functor<scan_t>
	Computes the generalized prefix sum over a simd value.
constexpr auto	eve::scatter = functor<scatter_t>
	Store a SIMD register to memory using scattered indexes.
constexpr auto	eve::sort = functor<sort_t>
	sorts a register in a accedning order accroding to a comparator.
constexpr auto	eve::zip = functor<zip_t>
	Callable for SoA value constructions.
constexpr callable_broadcast_group_	eve::broadcast_group = {}
	Computes the TODO.
constexpr callable_deinterleave_groups_shuffle_	eve::deinterleave_groups_shuffle = {}
	Callable object for a deinterleave groups shuffle.
constexpr splat_type const	eve::splat = {}
	Computes the TODO.