221 lines
10 KiB
Markdown
221 lines
10 KiB
Markdown
[simd.permute.memory]
|
||
|
||
# 29 Numerics library [[numerics]](./#numerics)
|
||
|
||
## 29.10 Data-parallel types [[simd]](simd#permute.memory)
|
||
|
||
### 29.10.8 basic_vec non-member operations [[simd.nonmembers]](simd.nonmembers#simd.permute.memory)
|
||
|
||
#### 29.10.8.10 simd memory permute [simd.permute.memory]
|
||
|
||
[ð](#lib:unchecked_gather_from,simd)
|
||
|
||
`template<class V = see below, ranges::[contiguous_range](range.refinements#concept:contiguous_range "25.4.6 Other range refinements [range.refinements]") R, [simd-integral](simd.expos#concept:simd-integral "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") I, class... Flags>
|
||
requires ranges::[sized_range](range.sized#concept:sized_range "25.4.4 Sized ranges [range.sized]")<R>
|
||
constexpr V unchecked_gather_from(R&& in, const I& indices, flags<Flags...> f = {});
|
||
template<class V = see below, ranges::[contiguous_range](range.refinements#concept:contiguous_range "25.4.6 Other range refinements [range.refinements]") R, [simd-integral](simd.expos#concept:simd-integral "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") I, class... Flags>
|
||
requires ranges::[sized_range](range.sized#concept:sized_range "25.4.4 Sized ranges [range.sized]")<R>
|
||
constexpr V unchecked_gather_from(R&& in, const typename I::mask_type& mask,
|
||
const I& indices, flags<Flags...> f = {});
|
||
`
|
||
|
||
[1](#1)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L18945)
|
||
|
||
Let mask be typename I::mask_type(true) for the overload with
|
||
no mask parameter[.](#1.sentence-1)
|
||
|
||
[2](#2)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L18949)
|
||
|
||
*Preconditions*: All values in select(mask, indices, typename I::value_type()) are in
|
||
the range [0, ranges::size(in))[.](#2.sentence-1)
|
||
|
||
[3](#3)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L18954)
|
||
|
||
*Effects*: Equivalent to: return partial_gather_from<V>(in, mask, indices, f);
|
||
|
||
[4](#4)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L18958)
|
||
|
||
*Remarks*: The default argument for template parameter V isvec<ranges::range_value_t<R>, I::size()>[.](#4.sentence-1)
|
||
|
||
[ð](#lib:partial_gather_from,simd)
|
||
|
||
`template<class V = see below, ranges::[contiguous_range](range.refinements#concept:contiguous_range "25.4.6 Other range refinements [range.refinements]") R, [simd-integral](simd.expos#concept:simd-integral "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") I, class... Flags>
|
||
requires ranges::[sized_range](range.sized#concept:sized_range "25.4.4 Sized ranges [range.sized]")<R>
|
||
constexpr V partial_gather_from(R&& in, const I& indices, flags<Flags...> f = {});
|
||
template<class V = see below, ranges::[contiguous_range](range.refinements#concept:contiguous_range "25.4.6 Other range refinements [range.refinements]") R, [simd-integral](simd.expos#concept:simd-integral "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") I, class... Flags>
|
||
requires ranges::[sized_range](range.sized#concept:sized_range "25.4.4 Sized ranges [range.sized]")<R>
|
||
constexpr V partial_gather_from(R&& in, const typename I::mask_type& mask,
|
||
const I& indices, flags<Flags...> f = {});
|
||
`
|
||
|
||
[5](#5)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L18976)
|
||
|
||
Let:
|
||
|
||
- [(5.1)](#5.1)
|
||
|
||
mask be typename I::mask_type(true) for the overload with no mask parameter;
|
||
|
||
- [(5.2)](#5.2)
|
||
|
||
T be typename V::value_type[.](#5.sentence-1)
|
||
|
||
[6](#6)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L18986)
|
||
|
||
*Mandates*:
|
||
|
||
- [(6.1)](#6.1)
|
||
|
||
ranges::range_value_t<R> is a vectorizable type,
|
||
|
||
- [(6.2)](#6.2)
|
||
|
||
same_as<remove_cvref_t<V>, V> is true,
|
||
|
||
- [(6.3)](#6.3)
|
||
|
||
V is an enabled specialization of basic_vec,
|
||
|
||
- [(6.4)](#6.4)
|
||
|
||
V::size() == I::size() is true, and
|
||
|
||
- [(6.5)](#6.5)
|
||
|
||
if the template parameter pack Flags does not contain*convert-flag*, then the conversion fromranges::range_value_t<R> to T is value-preserving[.](#6.sentence-1)
|
||
|
||
[7](#7)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19003)
|
||
|
||
*Preconditions*:
|
||
|
||
- [(7.1)](#7.1)
|
||
|
||
If the template parameter pack Flags contains*aligned-flag*, ranges::data(in) points to storage aligned byalignment_v<V, ranges::range_value_t<R>>[.](#7.1.sentence-1)
|
||
|
||
- [(7.2)](#7.2)
|
||
|
||
If the template parameter pack Flags contains*overaligned-flag*<N>, ranges::data(in) points to
|
||
storage aligned by N[.](#7.2.sentence-1)
|
||
|
||
[8](#8)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19016)
|
||
|
||
*Returns*: A basic_vec object where the ith element is initialized to the result ofmask[i] && indices[i] < ranges::size(in) ? static_cast<T>(ranges::data(in)[indices[i]]) : T() for all i in the range [0, I::size())[.](#8.sentence-1)
|
||
|
||
[9](#9)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19025)
|
||
|
||
*Remarks*: The default argument for template parameter V isvec<ranges::range_value_t<R>, I::size()>[.](#9.sentence-1)
|
||
|
||
[ð](#lib:unchecked_scatter_to,simd)
|
||
|
||
`template<[simd-vec-type](simd.expos#concept:simd-vec-type "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") V, ranges::[contiguous_range](range.refinements#concept:contiguous_range "25.4.6 Other range refinements [range.refinements]") R, [simd-integral](simd.expos#concept:simd-integral "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") I, class... Flags>
|
||
requires ranges::[sized_range](range.sized#concept:sized_range "25.4.4 Sized ranges [range.sized]")<R>
|
||
constexpr void unchecked_scatter_to(const V& v, R&& out, const I& indices,
|
||
flags<Flags...> f = {});
|
||
template<[simd-vec-type](simd.expos#concept:simd-vec-type "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") V, ranges::[contiguous_range](range.refinements#concept:contiguous_range "25.4.6 Other range refinements [range.refinements]") R, [simd-integral](simd.expos#concept:simd-integral "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") I, class... Flags>
|
||
requires ranges::[sized_range](range.sized#concept:sized_range "25.4.4 Sized ranges [range.sized]")<R>
|
||
constexpr void unchecked_scatter_to(const V& v, R&& out, const typename I::mask_type& mask,
|
||
const I& indices, flags<Flags...> f = {});
|
||
`
|
||
|
||
[10](#10)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19044)
|
||
|
||
Let mask be typename I::mask_type(true) for the overload with
|
||
no mask parameter[.](#10.sentence-1)
|
||
|
||
[11](#11)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19048)
|
||
|
||
*Preconditions*: All values in select(mask, indices, typename I::value_type()) are in
|
||
the range [0, ranges::size(out))[.](#11.sentence-1)
|
||
|
||
[12](#12)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19053)
|
||
|
||
*Effects*: Equivalent to: partial_scatter_to(v, out, mask, indices, f);
|
||
|
||
[ð](#lib:partial_scatter_to,simd)
|
||
|
||
`template<[simd-vec-type](simd.expos#concept:simd-vec-type "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") V, ranges::[contiguous_range](range.refinements#concept:contiguous_range "25.4.6 Other range refinements [range.refinements]") R, [simd-integral](simd.expos#concept:simd-integral "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") I, class... Flags>
|
||
requires ranges::[sized_range](range.sized#concept:sized_range "25.4.4 Sized ranges [range.sized]")<R>
|
||
constexpr void
|
||
partial_scatter_to(const V& v, R&& out, const I& indices, flags<Flags...> f = {});
|
||
template<[simd-vec-type](simd.expos#concept:simd-vec-type "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") V, ranges::[contiguous_range](range.refinements#concept:contiguous_range "25.4.6 Other range refinements [range.refinements]") R, [simd-integral](simd.expos#concept:simd-integral "29.10.2 Exposition-only types, variables, and concepts [simd.expos]") I, class... Flags>
|
||
requires ranges::[sized_range](range.sized#concept:sized_range "25.4.4 Sized ranges [range.sized]")<R>
|
||
constexpr void partial_scatter_to(const V& v, R&& out, const typename I::mask_type& mask,
|
||
const I& indices, flags<Flags...> f = {});
|
||
`
|
||
|
||
[13](#13)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19071)
|
||
|
||
Let mask be typename I::mask_type(true) for the overload with
|
||
no mask parameter[.](#13.sentence-1)
|
||
|
||
[14](#14)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19075)
|
||
|
||
*Constraints*: V::size() == I::size() is true[.](#14.sentence-1)
|
||
|
||
[15](#15)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19079)
|
||
|
||
*Mandates*:
|
||
|
||
- [(15.1)](#15.1)
|
||
|
||
ranges::range_value_t<R> is a vectorizable type, and
|
||
|
||
- [(15.2)](#15.2)
|
||
|
||
if the template parameter pack Flags does not contain *convert-flag*,
|
||
then the conversion from typename V::value_type to ranges::range_value_t<R> is value-preserving[.](#15.sentence-1)
|
||
|
||
[16](#16)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19090)
|
||
|
||
*Preconditions*:
|
||
|
||
- [(16.1)](#16.1)
|
||
|
||
For all selected indices i the values indices[i] are unique[.](#16.1.sentence-1)
|
||
|
||
- [(16.2)](#16.2)
|
||
|
||
If the template parameter pack Flags contains *aligned-flag*,ranges::data(out) points to storage aligned byalignment_v<V, ranges::range_value_t<R>>[.](#16.2.sentence-1)
|
||
|
||
- [(16.3)](#16.3)
|
||
|
||
If the template parameter pack Flags contains*overaligned-flag*<N>,ranges::data(out) points to storage aligned by N[.](#16.3.sentence-1)
|
||
|
||
[17](#17)
|
||
|
||
[#](http://github.com/Eelis/draft/tree/9adde4bc1c62ec234483e63ea3b70a59724c745a/source/numerics.tex#L19105)
|
||
|
||
*Effects*: For all i in the range [0, I::size()), if mask[i] &&(indices[i] < ranges::size(out)) is true, evaluatesranges::data(out)[indices[i]] = v[i][.](#17.sentence-1)
|