#[repr(simd)]pub struct Simd<T, const N: usize>(/* private fields */)
where
LaneCount<N>: SupportedLaneCount,
T: SimdElement;
portable_simd
#86656)Expand description
A SIMD vector with the shape of [T; N]
but the operations of T
.
Simd<T, N>
supports the operators (+, *, etc.) that T
does in “elementwise” fashion.
These take the element at each index from the left-hand side and right-hand side,
perform the operation, then return the result in the same index in a vector of equal size.
However, Simd
differs from normal iteration and normal arrays:
Simd<T, N>
executesN
operations in a single step with nobreak
sSimd<T, N>
can have an alignment greater thanT
, for better mechanical sympathy
By always imposing these constraints on Simd
, it is easier to compile elementwise operations
into machine instructions that can themselves be executed in parallel.
let a: [i32; 4] = [-2, 0, 2, 4];
let b = [10, 9, 8, 7];
let sum = array::from_fn(|i| a[i] + b[i]);
let prod = array::from_fn(|i| a[i] * b[i]);
// `Simd<T, N>` implements `From<[T; N]>`
let (v, w) = (Simd::from(a), Simd::from(b));
// Which means arrays implement `Into<Simd<T, N>>`.
assert_eq!(v + w, sum.into());
assert_eq!(v * w, prod.into());
Simd
with integer elements treats operators as wrapping, as if T
was Wrapping<T>
.
Thus, Simd
does not implement wrapping_add
, because that is the default behavior.
This means there is no warning on overflows, even in “debug” builds.
For most applications where Simd
is appropriate, it is “not a bug” to wrap,
and even “debug builds” are unlikely to tolerate the loss of performance.
You may want to consider using explicitly checked arithmetic if such is required.
Division by zero on integers still causes a panic, so
you may want to consider using f32
or f64
if that is unacceptable.
§Layout
Simd<T, N>
has a layout similar to [T; N]
(identical “shapes”), with a greater alignment.
[T; N]
is aligned to T
, but Simd<T, N>
will have an alignment based on both T
and N
.
Thus it is sound to transmute
Simd<T, N>
to [T; N]
and should optimize to “zero cost”,
but the reverse transmutation may require a copy the compiler cannot simply elide.
§ABI “Features”
Due to Rust’s safety guarantees, Simd<T, N>
is currently passed and returned via memory,
not SIMD registers, except as an optimization. Using #[inline]
on functions that accept
Simd<T, N>
or return it is recommended, at the cost of code generation time, as
inlining SIMD-using functions can omit a large function prolog or epilog and thus
improve both speed and code size. The need for this may be corrected in the future.
Using #[inline(always)]
still requires additional care.
§Safe SIMD with Unsafe Rust
Operations with Simd
are typically safe, but there are many reasons to want to combine SIMD with unsafe
code.
Care must be taken to respect differences between Simd
and other types it may be transformed into or derived from.
In particular, the layout of Simd<T, N>
may be similar to [T; N]
, and may allow some transmutations,
but references to [T; N]
are not interchangeable with those to Simd<T, N>
.
Thus, when using unsafe
Rust to read and write Simd<T, N>
through raw pointers, it is a good idea to first try with
read_unaligned
and write_unaligned
. This is because:
read
andwrite
require full alignment (in this case,Simd<T, N>
’s alignment)Simd<T, N>
is often read from or written to[T]
and other types aligned toT
- combining these actions violates the
unsafe
contract and explodes the program into a puff of undefined behavior - the compiler can implicitly adjust layouts to make unaligned reads or writes fully aligned if it sees the optimization
- most contemporary processors with “aligned” and “unaligned” read and write instructions exhibit no performance difference if the “unaligned” variant is aligned at runtime
Less obligations mean unaligned reads and writes are less likely to make the program unsound,
and may be just as fast as stricter alternatives.
When trying to guarantee alignment, [T]::as_simd
is an option for
converting [T]
to [Simd<T, N>]
, and allows soundly operating on an aligned SIMD body,
but it may cost more time when handling the scalar head and tail.
If these are not enough, it is most ideal to design data structures to be already aligned
to mem::align_of::<Simd<T, N>>()
before using unsafe
Rust to read or write.
Other ways to compensate for these facts, like materializing Simd
to or from an array first,
are handled by safe methods like Simd::from_array
and Simd::from_slice
.
Implementations§
source§impl<T, const N: usize> Simd<T, N>
impl<T, const N: usize> Simd<T, N>
sourcepub fn reverse(self) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn reverse(self) -> Simd<T, N>
portable_simd
#86656)Reverse the order of the elements in the vector.
sourcepub fn rotate_elements_left<const OFFSET: usize>(self) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn rotate_elements_left<const OFFSET: usize>(self) -> Simd<T, N>
portable_simd
#86656)Rotates the vector such that the first OFFSET
elements of the slice move to the end
while the last self.len() - OFFSET
elements move to the front. After calling rotate_elements_left
,
the element previously at index OFFSET
will become the first element in the slice.
sourcepub fn rotate_elements_right<const OFFSET: usize>(self) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn rotate_elements_right<const OFFSET: usize>(self) -> Simd<T, N>
portable_simd
#86656)Rotates the vector such that the first self.len() - OFFSET
elements of the vector move to
the end while the last OFFSET
elements move to the front. After calling rotate_elements_right
,
the element previously at index self.len() - OFFSET
will become the first element in the slice.
sourcepub fn interleave(self, other: Simd<T, N>) -> (Simd<T, N>, Simd<T, N>)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn interleave(self, other: Simd<T, N>) -> (Simd<T, N>, Simd<T, N>)
portable_simd
#86656)Interleave two vectors.
The resulting vectors contain elements taken alternatively from self
and other
, first
filling the first result, and then the second.
The reverse of this operation is Simd::deinterleave
.
sourcepub fn deinterleave(self, other: Simd<T, N>) -> (Simd<T, N>, Simd<T, N>)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn deinterleave(self, other: Simd<T, N>) -> (Simd<T, N>, Simd<T, N>)
portable_simd
#86656)Deinterleave two vectors.
The first result takes every other element of self
and then other
, starting with
the first element.
The second result takes every other element of self
and then other
, starting with
the second element.
The reverse of this operation is Simd::interleave
.
sourcepub fn resize<const M: usize>(self, value: T) -> Simd<T, M>where
LaneCount<M>: SupportedLaneCount,
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn resize<const M: usize>(self, value: T) -> Simd<T, M>where
LaneCount<M>: SupportedLaneCount,
portable_simd
#86656)Resize a vector.
If M
> N
, extends the length of a vector, setting the new elements to value
.
If M
< N
, truncates the vector to the first M
elements.
source§impl<const N: usize> Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
impl<const N: usize> Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
sourcepub fn swizzle_dyn(self, idxs: Simd<u8, N>) -> Simd<u8, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn swizzle_dyn(self, idxs: Simd<u8, N>) -> Simd<u8, N>
portable_simd
#86656)Swizzle a vector of bytes according to the index vector. Indices within range select the appropriate byte. Indices “out of bounds” instead select 0.
Note that the current implementation is selected during build-time
of the standard library, so cargo build -Zbuild-std
may be necessary
to unlock better performance, especially for larger vectors.
A planned compiler improvement will enable using #[target_feature]
instead.
source§impl<T, const N: usize> Simd<T, N>
impl<T, const N: usize> Simd<T, N>
sourcepub const LEN: usize = N
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub const LEN: usize = N
portable_simd
#86656)Number of elements in this vector.
sourcepub const fn len(&self) -> usize
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub const fn len(&self) -> usize
portable_simd
#86656)Returns the number of elements in this SIMD vector.
§Examples
sourcepub fn splat(value: T) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn splat(value: T) -> Simd<T, N>
portable_simd
#86656)Constructs a new SIMD vector with all elements set to the given value.
§Examples
sourcepub const fn as_array(&self) -> &[T; N]
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub const fn as_array(&self) -> &[T; N]
portable_simd
#86656)Returns an array reference containing the entire SIMD vector.
§Examples
sourcepub fn as_mut_array(&mut self) -> &mut [T; N]
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn as_mut_array(&mut self) -> &mut [T; N]
portable_simd
#86656)Returns a mutable array reference containing the entire SIMD vector.
sourcepub const fn from_array(array: [T; N]) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub const fn from_array(array: [T; N]) -> Simd<T, N>
portable_simd
#86656)Converts an array to a SIMD vector.
sourcepub const fn to_array(self) -> [T; N]
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub const fn to_array(self) -> [T; N]
portable_simd
#86656)Converts a SIMD vector to an array.
sourcepub const fn from_slice(slice: &[T]) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub const fn from_slice(slice: &[T]) -> Simd<T, N>
portable_simd
#86656)sourcepub fn copy_to_slice(self, slice: &mut [T])
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn copy_to_slice(self, slice: &mut [T])
portable_simd
#86656)sourcepub fn load_or_default(slice: &[T]) -> Simd<T, N>where
T: Default,
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn load_or_default(slice: &[T]) -> Simd<T, N>where
T: Default,
portable_simd
#86656)Reads contiguous elements from slice
. Elements are read so long as they’re in-bounds for
the slice
. Otherwise, the default value for the element type is returned.
§Examples
sourcepub fn load_or(slice: &[T], or: Simd<T, N>) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn load_or(slice: &[T], or: Simd<T, N>) -> Simd<T, N>
portable_simd
#86656)Reads contiguous elements from slice
. Elements are read so long as they’re in-bounds for
the slice
. Otherwise, the corresponding value from or
is passed through.
§Examples
sourcepub fn load_select_or_default(
slice: &[T],
enable: Mask<<T as SimdElement>::Mask, N>,
) -> Simd<T, N>where
T: Default,
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn load_select_or_default(
slice: &[T],
enable: Mask<<T as SimdElement>::Mask, N>,
) -> Simd<T, N>where
T: Default,
portable_simd
#86656)Reads contiguous elements from slice
. Each element is read from memory if its
corresponding element in enable
is true
.
When the element is disabled or out of bounds for the slice, that memory location
is not accessed and the corresponding value from or
is passed through.
§Examples
sourcepub fn load_select(
slice: &[T],
enable: Mask<<T as SimdElement>::Mask, N>,
or: Simd<T, N>,
) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn load_select( slice: &[T], enable: Mask<<T as SimdElement>::Mask, N>, or: Simd<T, N>, ) -> Simd<T, N>
portable_simd
#86656)Reads contiguous elements from slice
. Each element is read from memory if its
corresponding element in enable
is true
.
When the element is disabled or out of bounds for the slice, that memory location
is not accessed and the corresponding value from or
is passed through.
§Examples
sourcepub unsafe fn load_select_unchecked(
slice: &[T],
enable: Mask<<T as SimdElement>::Mask, N>,
or: Simd<T, N>,
) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn load_select_unchecked( slice: &[T], enable: Mask<<T as SimdElement>::Mask, N>, or: Simd<T, N>, ) -> Simd<T, N>
portable_simd
#86656)Reads contiguous elements from slice
. Each element is read from memory if its
corresponding element in enable
is true
.
When the element is disabled, that memory location is not accessed and the corresponding
value from or
is passed through.
sourcepub unsafe fn load_select_ptr(
ptr: *const T,
enable: Mask<<T as SimdElement>::Mask, N>,
or: Simd<T, N>,
) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn load_select_ptr( ptr: *const T, enable: Mask<<T as SimdElement>::Mask, N>, or: Simd<T, N>, ) -> Simd<T, N>
portable_simd
#86656)Reads contiguous elements starting at ptr
. Each element is read from memory if its
corresponding element in enable
is true
.
When the element is disabled, that memory location is not accessed and the corresponding
value from or
is passed through.
sourcepub fn gather_or(
slice: &[T],
idxs: Simd<usize, N>,
or: Simd<T, N>,
) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn gather_or( slice: &[T], idxs: Simd<usize, N>, or: Simd<T, N>, ) -> Simd<T, N>
portable_simd
#86656)Reads from potentially discontiguous indices in slice
to construct a SIMD vector.
If an index is out-of-bounds, the element is instead selected from the or
vector.
§Examples
sourcepub fn gather_or_default(slice: &[T], idxs: Simd<usize, N>) -> Simd<T, N>where
T: Default,
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn gather_or_default(slice: &[T], idxs: Simd<usize, N>) -> Simd<T, N>where
T: Default,
portable_simd
#86656)Reads from indices in slice
to construct a SIMD vector.
If an index is out-of-bounds, the element is set to the default given by T: Default
.
§Examples
sourcepub fn gather_select(
slice: &[T],
enable: Mask<isize, N>,
idxs: Simd<usize, N>,
or: Simd<T, N>,
) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn gather_select( slice: &[T], enable: Mask<isize, N>, idxs: Simd<usize, N>, or: Simd<T, N>, ) -> Simd<T, N>
portable_simd
#86656)Reads from indices in slice
to construct a SIMD vector.
The mask enable
s all true
indices and disables all false
indices.
If an index is disabled or is out-of-bounds, the element is selected from the or
vector.
§Examples
let vec: Vec<i32> = vec![10, 11, 12, 13, 14, 15, 16, 17, 18];
let idxs = Simd::from_array([9, 3, 0, 5]); // Includes an out-of-bounds index
let alt = Simd::from_array([-5, -4, -3, -2]);
let enable = Mask::from_array([true, true, true, false]); // Includes a masked element
let result = Simd::gather_select(&vec, enable, idxs, alt);
assert_eq!(result, Simd::from_array([-5, 13, 10, -2]));
sourcepub unsafe fn gather_select_unchecked(
slice: &[T],
enable: Mask<isize, N>,
idxs: Simd<usize, N>,
or: Simd<T, N>,
) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn gather_select_unchecked( slice: &[T], enable: Mask<isize, N>, idxs: Simd<usize, N>, or: Simd<T, N>, ) -> Simd<T, N>
portable_simd
#86656)Reads from indices in slice
to construct a SIMD vector.
The mask enable
s all true
indices and disables all false
indices.
If an index is disabled, the element is selected from the or
vector.
§Safety
Calling this function with an enable
d out-of-bounds index is undefined behavior
even if the resulting value is not used.
§Examples
let vec: Vec<i32> = vec![10, 11, 12, 13, 14, 15, 16, 17, 18];
let idxs = Simd::from_array([9, 3, 0, 5]); // Includes an out-of-bounds index
let alt = Simd::from_array([-5, -4, -3, -2]);
let enable = Mask::from_array([true, true, true, false]); // Includes a masked element
// If this mask was used to gather, it would be unsound. Let's fix that.
let enable = enable & idxs.simd_lt(Simd::splat(vec.len()));
// The out-of-bounds index has been masked, so it's safe to gather now.
let result = unsafe { Simd::gather_select_unchecked(&vec, enable, idxs, alt) };
assert_eq!(result, Simd::from_array([-5, 13, 10, -2]));
sourcepub unsafe fn gather_ptr(source: Simd<*const T, N>) -> Simd<T, N>where
T: Default,
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn gather_ptr(source: Simd<*const T, N>) -> Simd<T, N>where
T: Default,
portable_simd
#86656)Reads elementwise from pointers into a SIMD vector.
§Safety
Each read must satisfy the same conditions as core::ptr::read
.
§Example
sourcepub unsafe fn gather_select_ptr(
source: Simd<*const T, N>,
enable: Mask<isize, N>,
or: Simd<T, N>,
) -> Simd<T, N>
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn gather_select_ptr( source: Simd<*const T, N>, enable: Mask<isize, N>, or: Simd<T, N>, ) -> Simd<T, N>
portable_simd
#86656)Conditionally read elementwise from pointers into a SIMD vector.
The mask enable
s all true
pointers and disables all false
pointers.
If a pointer is disabled, the element is selected from the or
vector,
and no read is performed.
§Safety
Enabled elements must satisfy the same conditions as core::ptr::read
.
§Example
let values = [6, 2, 4, 9];
let enable = Mask::from_array([true, true, false, true]);
let offsets = Simd::from_array([1, 0, 0, 3]);
let source = Simd::splat(values.as_ptr()).wrapping_add(offsets);
let gathered = unsafe { Simd::gather_select_ptr(source, enable, Simd::splat(0)) };
assert_eq!(gathered, Simd::from_array([2, 6, 0, 9]));
sourcepub fn store_select(
self,
slice: &mut [T],
enable: Mask<<T as SimdElement>::Mask, N>,
)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn store_select( self, slice: &mut [T], enable: Mask<<T as SimdElement>::Mask, N>, )
portable_simd
#86656)Conditionally write contiguous elements to slice
. The enable
mask controls
which elements are written, as long as they’re in-bounds of the slice
.
If the element is disabled or out of bounds, no memory access to that location
is made.
§Examples
sourcepub unsafe fn store_select_unchecked(
self,
slice: &mut [T],
enable: Mask<<T as SimdElement>::Mask, N>,
)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn store_select_unchecked( self, slice: &mut [T], enable: Mask<<T as SimdElement>::Mask, N>, )
portable_simd
#86656)sourcepub unsafe fn store_select_ptr(
self,
ptr: *mut T,
enable: Mask<<T as SimdElement>::Mask, N>,
)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn store_select_ptr( self, ptr: *mut T, enable: Mask<<T as SimdElement>::Mask, N>, )
portable_simd
#86656)Conditionally write contiguous elements starting from ptr
.
The enable
mask controls which elements are written.
When disabled, the memory location corresponding to that element is not accessed.
§Safety
Memory addresses for element are calculated pointer::wrapping_offset
and
each enabled element must satisfy the same conditions as core::ptr::write
.
sourcepub fn scatter(self, slice: &mut [T], idxs: Simd<usize, N>)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn scatter(self, slice: &mut [T], idxs: Simd<usize, N>)
portable_simd
#86656)Writes the values in a SIMD vector to potentially discontiguous indices in slice
.
If an index is out-of-bounds, the write is suppressed without panicking.
If two elements in the scattered vector would write to the same index
only the last element is guaranteed to actually be written.
§Examples
let mut vec: Vec<i32> = vec![10, 11, 12, 13, 14, 15, 16, 17, 18];
let idxs = Simd::from_array([9, 3, 0, 0]); // Note the duplicate index.
let vals = Simd::from_array([-27, 82, -41, 124]);
vals.scatter(&mut vec, idxs); // two logical writes means the last wins.
assert_eq!(vec, vec![124, 11, 12, 82, 14, 15, 16, 17, 18]);
sourcepub fn scatter_select(
self,
slice: &mut [T],
enable: Mask<isize, N>,
idxs: Simd<usize, N>,
)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub fn scatter_select( self, slice: &mut [T], enable: Mask<isize, N>, idxs: Simd<usize, N>, )
portable_simd
#86656)Writes values from a SIMD vector to multiple potentially discontiguous indices in slice
.
The mask enable
s all true
indices and disables all false
indices.
If an enabled index is out-of-bounds, the write is suppressed without panicking.
If two enabled elements in the scattered vector would write to the same index,
only the last element is guaranteed to actually be written.
§Examples
let mut vec: Vec<i32> = vec![10, 11, 12, 13, 14, 15, 16, 17, 18];
let idxs = Simd::from_array([9, 3, 0, 0]); // Includes an out-of-bounds index
let vals = Simd::from_array([-27, 82, -41, 124]);
let enable = Mask::from_array([true, true, true, false]); // Includes a masked element
vals.scatter_select(&mut vec, enable, idxs); // The last write is masked, thus omitted.
assert_eq!(vec, vec![-41, 11, 12, 82, 14, 15, 16, 17, 18]);
sourcepub unsafe fn scatter_select_unchecked(
self,
slice: &mut [T],
enable: Mask<isize, N>,
idxs: Simd<usize, N>,
)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn scatter_select_unchecked( self, slice: &mut [T], enable: Mask<isize, N>, idxs: Simd<usize, N>, )
portable_simd
#86656)Writes values from a SIMD vector to multiple potentially discontiguous indices in slice
.
The mask enable
s all true
indices and disables all false
indices.
If two enabled elements in the scattered vector would write to the same index,
only the last element is guaranteed to actually be written.
§Safety
Calling this function with an enabled out-of-bounds index is undefined behavior, and may lead to memory corruption.
§Examples
let mut vec: Vec<i32> = vec![10, 11, 12, 13, 14, 15, 16, 17, 18];
let idxs = Simd::from_array([9, 3, 0, 0]);
let vals = Simd::from_array([-27, 82, -41, 124]);
let enable = Mask::from_array([true, true, true, false]); // Masks the final index
// If this mask was used to scatter, it would be unsound. Let's fix that.
let enable = enable & idxs.simd_lt(Simd::splat(vec.len()));
// We have masked the OOB index, so it's safe to scatter now.
unsafe { vals.scatter_select_unchecked(&mut vec, enable, idxs); }
// The second write to index 0 was masked, thus omitted.
assert_eq!(vec, vec![-41, 11, 12, 82, 14, 15, 16, 17, 18]);
sourcepub unsafe fn scatter_ptr(self, dest: Simd<*mut T, N>)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn scatter_ptr(self, dest: Simd<*mut T, N>)
portable_simd
#86656)Writes pointers elementwise into a SIMD vector.
§Safety
Each write must satisfy the same conditions as core::ptr::write
.
§Example
sourcepub unsafe fn scatter_select_ptr(
self,
dest: Simd<*mut T, N>,
enable: Mask<isize, N>,
)
🔬This is a nightly-only experimental API. (portable_simd
#86656)
pub unsafe fn scatter_select_ptr( self, dest: Simd<*mut T, N>, enable: Mask<isize, N>, )
portable_simd
#86656)Conditionally write pointers elementwise into a SIMD vector.
The mask enable
s all true
pointers and disables all false
pointers.
If a pointer is disabled, the write to its pointee is skipped.
§Safety
Enabled pointers must satisfy the same conditions as core::ptr::write
.
§Example
let mut values = [0; 4];
let offset = Simd::from_array([3, 2, 1, 0]);
let ptrs = Simd::splat(values.as_mut_ptr()).wrapping_add(offset);
let enable = Mask::from_array([true, true, false, false]);
unsafe { Simd::from_array([6, 3, 5, 7]).scatter_select_ptr(ptrs, enable); }
assert_eq!(values, [0, 0, 3, 6]);
Trait Implementations§
source§impl<T, U, const N: usize> AddAssign<U> for Simd<T, N>
impl<T, U, const N: usize> AddAssign<U> for Simd<T, N>
source§fn add_assign(&mut self, rhs: U)
fn add_assign(&mut self, rhs: U)
+=
operation. Read moresource§impl<T, U, const N: usize> BitAndAssign<U> for Simd<T, N>
impl<T, U, const N: usize> BitAndAssign<U> for Simd<T, N>
source§fn bitand_assign(&mut self, rhs: U)
fn bitand_assign(&mut self, rhs: U)
&=
operation. Read moresource§impl<T, U, const N: usize> BitOrAssign<U> for Simd<T, N>
impl<T, U, const N: usize> BitOrAssign<U> for Simd<T, N>
source§fn bitor_assign(&mut self, rhs: U)
fn bitor_assign(&mut self, rhs: U)
|=
operation. Read moresource§impl<T, U, const N: usize> BitXorAssign<U> for Simd<T, N>
impl<T, U, const N: usize> BitXorAssign<U> for Simd<T, N>
source§fn bitxor_assign(&mut self, rhs: U)
fn bitxor_assign(&mut self, rhs: U)
^=
operation. Read moresource§impl<T, U, const N: usize> DivAssign<U> for Simd<T, N>
impl<T, U, const N: usize> DivAssign<U> for Simd<T, N>
source§fn div_assign(&mut self, rhs: U)
fn div_assign(&mut self, rhs: U)
/=
operation. Read moresource§impl<T, U, const N: usize> MulAssign<U> for Simd<T, N>
impl<T, U, const N: usize> MulAssign<U> for Simd<T, N>
source§fn mul_assign(&mut self, rhs: U)
fn mul_assign(&mut self, rhs: U)
*=
operation. Read moresource§impl<T, const N: usize> Ord for Simd<T, N>
impl<T, const N: usize> Ord for Simd<T, N>
1.21.0 · source§fn max(self, other: Self) -> Selfwhere
Self: Sized,
fn max(self, other: Self) -> Selfwhere
Self: Sized,
source§impl<T, const N: usize> PartialOrd for Simd<T, N>
impl<T, const N: usize> PartialOrd for Simd<T, N>
source§impl<'a, const N: usize> Product<&'a Simd<f32, N>> for Simd<f32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<f32, N>> for Simd<f32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<f64, N>> for Simd<f64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<f64, N>> for Simd<f64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<i16, N>> for Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<i16, N>> for Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<i32, N>> for Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<i32, N>> for Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<i64, N>> for Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<i64, N>> for Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<i8, N>> for Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<i8, N>> for Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<isize, N>> for Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<isize, N>> for Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<u16, N>> for Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<u16, N>> for Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<u32, N>> for Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<u32, N>> for Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<u64, N>> for Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<u64, N>> for Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<u8, N>> for Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<u8, N>> for Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'a, const N: usize> Product<&'a Simd<usize, N>> for Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'a, const N: usize> Product<&'a Simd<usize, N>> for Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<T, U, const N: usize> RemAssign<U> for Simd<T, N>
impl<T, U, const N: usize> RemAssign<U> for Simd<T, N>
source§fn rem_assign(&mut self, rhs: U)
fn rem_assign(&mut self, rhs: U)
%=
operation. Read moresource§impl<'lhs, const N: usize> Shl<&i16> for &'lhs Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&i16> for &'lhs Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&i32> for &'lhs Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&i32> for &'lhs Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&i64> for &'lhs Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&i64> for &'lhs Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&i8> for &'lhs Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&i8> for &'lhs Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&isize> for &'lhs Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&isize> for &'lhs Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&u16> for &'lhs Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&u16> for &'lhs Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&u32> for &'lhs Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&u32> for &'lhs Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&u64> for &'lhs Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&u64> for &'lhs Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&u8> for &'lhs Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&u8> for &'lhs Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<&usize> for &'lhs Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<&usize> for &'lhs Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<i16> for &'lhs Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<i16> for &'lhs Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<i32> for &'lhs Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<i32> for &'lhs Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<i64> for &'lhs Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<i64> for &'lhs Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<i8> for &'lhs Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<i8> for &'lhs Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<isize> for &'lhs Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<isize> for &'lhs Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<u16> for &'lhs Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<u16> for &'lhs Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<u32> for &'lhs Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<u32> for &'lhs Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<u64> for &'lhs Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<u64> for &'lhs Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<u8> for &'lhs Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<u8> for &'lhs Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shl<usize> for &'lhs Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shl<usize> for &'lhs Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<T, U, const N: usize> ShlAssign<U> for Simd<T, N>
impl<T, U, const N: usize> ShlAssign<U> for Simd<T, N>
source§fn shl_assign(&mut self, rhs: U)
fn shl_assign(&mut self, rhs: U)
<<=
operation. Read moresource§impl<'lhs, const N: usize> Shr<&i16> for &'lhs Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&i16> for &'lhs Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&i32> for &'lhs Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&i32> for &'lhs Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&i64> for &'lhs Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&i64> for &'lhs Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&i8> for &'lhs Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&i8> for &'lhs Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&isize> for &'lhs Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&isize> for &'lhs Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&u16> for &'lhs Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&u16> for &'lhs Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&u32> for &'lhs Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&u32> for &'lhs Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&u64> for &'lhs Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&u64> for &'lhs Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&u8> for &'lhs Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&u8> for &'lhs Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<&usize> for &'lhs Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<&usize> for &'lhs Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<i16> for &'lhs Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<i16> for &'lhs Simd<i16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<i32> for &'lhs Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<i32> for &'lhs Simd<i32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<i64> for &'lhs Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<i64> for &'lhs Simd<i64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<i8> for &'lhs Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<i8> for &'lhs Simd<i8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<isize> for &'lhs Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<isize> for &'lhs Simd<isize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<u16> for &'lhs Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<u16> for &'lhs Simd<u16, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<u32> for &'lhs Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<u32> for &'lhs Simd<u32, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<u64> for &'lhs Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<u64> for &'lhs Simd<u64, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<u8> for &'lhs Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<u8> for &'lhs Simd<u8, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<'lhs, const N: usize> Shr<usize> for &'lhs Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
impl<'lhs, const N: usize> Shr<usize> for &'lhs Simd<usize, N>where
LaneCount<N>: SupportedLaneCount,
source§impl<T, U, const N: usize> ShrAssign<U> for Simd<T, N>
impl<T, U, const N: usize> ShrAssign<U> for Simd<T, N>
source§fn shr_assign(&mut self, rhs: U)
fn shr_assign(&mut self, rhs: U)
>>=
operation. Read moresource§impl<T, const N: usize> SimdConstPtr for Simd<*const T, N>where
LaneCount<N>: SupportedLaneCount,
impl<T, const N: usize> SimdConstPtr for Simd<*const T, N>where
LaneCount<N>: SupportedLaneCount,
source§type Usize = Simd<usize, N>
type Usize = Simd<usize, N>
portable_simd
#86656)usize
with the same number of elements.source§type Isize = Simd<isize, N>
type Isize = Simd<isize, N>
portable_simd
#86656)isize
with the same number of elements.source§type CastPtr<U> = Simd<*const U, N>
type CastPtr<U> = Simd<*const U, N>
portable_simd
#86656)source§type MutPtr = Simd<*mut T, N>
type MutPtr = Simd<*mut T, N>
portable_simd
#86656)source§type Mask = Mask<isize, N>
type Mask = Mask<isize, N>
portable_simd
#86656)source§fn is_null(self) -> <Simd<*const T, N> as SimdConstPtr>::Mask
fn is_null(self) -> <Simd<*const T, N> as SimdConstPtr>::Mask
portable_simd
#86656)true
for each element that is null.source§fn cast<U>(self) -> <Simd<*const T, N> as SimdConstPtr>::CastPtr<U>
fn cast<U>(self) -> <Simd<*const T, N> as SimdConstPtr>::CastPtr<U>
portable_simd
#86656)source§fn cast_mut(self) -> <Simd<*const T, N> as SimdConstPtr>::MutPtr
fn cast_mut(self) -> <Simd<*const T, N> as SimdConstPtr>::MutPtr
portable_simd
#86656)source§fn addr(self) -> <Simd<*const T, N> as SimdConstPtr>::Usize
fn addr(self) -> <Simd<*const T, N> as SimdConstPtr>::Usize
portable_simd
#86656)source§fn with_addr(
self,
addr: <Simd<*const T, N> as SimdConstPtr>::Usize,
) -> Simd<*const T, N>
fn with_addr( self, addr: <Simd<*const T, N> as SimdConstPtr>::Usize, ) -> Simd<*const T, N>
portable_simd
#86656)source§fn expose_provenance(self) -> <Simd<*const T, N> as SimdConstPtr>::Usize
fn expose_provenance(self) -> <Simd<*const T, N> as SimdConstPtr>::Usize
portable_simd
#86656)Self::with_exposed_provenance
and returns the “address” portion.source§fn with_exposed_provenance(
addr: <Simd<*const T, N> as SimdConstPtr>::Usize,
) -> Simd<*const T, N>
fn with_exposed_provenance( addr: <Simd<*const T, N> as SimdConstPtr>::Usize, ) -> Simd<*const T, N>
portable_simd
#86656)source§fn wrapping_offset(
self,
count: <Simd<*const T, N> as SimdConstPtr>::Isize,
) -> Simd<*const T, N>
fn wrapping_offset( self, count: <Simd<*const T, N> as SimdConstPtr>::Isize, ) -> Simd<*const T, N>
portable_simd
#86656)source§impl<const N: usize> SimdFloat for Simd<f32, N>where
LaneCount<N>: SupportedLaneCount,
impl<const N: usize> SimdFloat for Simd<f32, N>where
LaneCount<N>: SupportedLaneCount,
source§type Mask = Mask<<i32 as SimdElement>::Mask, N>
type Mask = Mask<<i32 as SimdElement>::Mask, N>
portable_simd
#86656)source§type Scalar = f32
type Scalar = f32
portable_simd
#86656)source§type Bits = Simd<u32, N>
type Bits = Simd<u32, N>
portable_simd
#86656)