wide

wide copied to clipboard

Boolean masks should become their own type

3

some people like the increased type safety. this is a breaking change so it'd be in 0.6, but otherwise it's not hard.

Lokathor

Give i32x4 a From impl that uses _mm_set1_epi32

1

The `impl From for i32x4` uses new(i32, i32, i32, i32) always. There's an intrinsic to not do that. ``` impl From for i32x4 { #[inline] #[must_use] fn from(i: i32) ->...

mooman219

NEON instructions on `aarch64`?

4

I know the README says this: > ... and on other architectures this is done by carefully writing functions so that LLVM hopefully does the right thing But what magic...

torokati44

RRRadicalEdward

Add `from_le_bytes` and `to_le_bytes` for SIMD types

1

suhr

Add hyperbolic geometric functions such as `sinh`

1

At the moment hyperbolic geometric functions are missing. Can these be added, or are they not implemented as SIMD instructions on all platforms?

benruijl

UnrealInReal

Feature idea/request: `clamp` and `signum`

2

They came up in an algorithm I'm considering for SIMDification (https://github.com/ruffle-rs/h263-rs/pull/49), and perhaps they would came in handy for others as well - they are in `f32` at least, after...

torokati44

512-bit types and AVX-512 support

1

This crate seems to lack 512-bit types and AVX-512 support, which is becoming quite important..

kytans

Add fast type conversion methods where possible

10

For SIMD types that have intrinsics for quickly converting to another type, methods were added to leverage these intrinsics for fast conversion.

shssoichiro