Skip to content

Add interleave method #206

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
LaurenzV merged 15 commits into from
Apr 18, 2026
Merged

Add interleave method #206

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
15 commits
Select commit Hold shift + click to select a range
9c754c6
Add interleave function mimicking std::simd
Shnatsel Jan 21, 2026
f5dac39
Merge branch 'main' into interleave
Shnatsel Apr 12, 2026
0f0f905
Add an optimized AVX2 implementation of interleave
Shnatsel Apr 12, 2026
f689305
re-run the generator
Shnatsel Apr 12, 2026
b8e4adc
cargo fmt
Shnatsel Apr 12, 2026
2abf59e
Add basic tests for deinterleaving
Shnatsel Apr 12, 2026
5c87962
Implement deinterleave() operation in the generator
Shnatsel Apr 12, 2026
f058be3
Re-run the generator
Shnatsel Apr 12, 2026
7b59999
Add more interleave/deinterleave tests
Shnatsel Apr 12, 2026
c57f5e2
Document that interleave/deinterleave are faster than zip_low followe…
Shnatsel Apr 12, 2026
7642be3
re-run the generator
Shnatsel Apr 12, 2026
b767998
Reduce comments
LaurenzV Apr 18, 2026
5889e19
Reduce comments
LaurenzV Apr 18, 2026
aaa0cac
Reformat
LaurenzV Apr 18, 2026
781c7fe
Remove more comments
LaurenzV Apr 18, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
392 changes: 392 additions & 0 deletions fearless_simd/src/generated/avx2.rs
View file
Open in desktop

Large diffs are not rendered by default.

480 changes: 480 additions & 0 deletions fearless_simd/src/generated/fallback.rs
View file
Open in desktop

Large diffs are not rendered by default.

480 changes: 480 additions & 0 deletions fearless_simd/src/generated/neon.rs
View file
Open in desktop

Large diffs are not rendered by default.

312 changes: 208 additions & 104 deletions fearless_simd/src/generated/simd_trait.rs
View file
Open in desktop

Large diffs are not rendered by default.

199 changes: 199 additions & 0 deletions fearless_simd/src/generated/simd_types.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -177,6 +177,14 @@ impl<S: Simd> crate::SimdFloat<S> for f32x4<S> {
self.simd.unzip_high_f32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_f32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_f32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn max(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.max_f32x4(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -405,6 +413,14 @@ impl<S: Simd> crate::SimdInt<S> for i8x16<S> {
self.simd.unzip_high_i8x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i8x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_i8x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i8x16(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -581,6 +597,14 @@ impl<S: Simd> crate::SimdInt<S> for u8x16<S> {
self.simd.unzip_high_u8x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u8x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_u8x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u8x16(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -914,6 +938,14 @@ impl<S: Simd> crate::SimdInt<S> for i16x8<S> {
self.simd.unzip_high_i16x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i16x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_i16x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i16x8(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -1090,6 +1122,14 @@ impl<S: Simd> crate::SimdInt<S> for u16x8<S> {
self.simd.unzip_high_u16x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u16x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_u16x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u16x8(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -1423,6 +1463,14 @@ impl<S: Simd> crate::SimdInt<S> for i32x4<S> {
self.simd.unzip_high_i32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_i32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i32x4(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -1611,6 +1659,14 @@ impl<S: Simd> crate::SimdInt<S> for u32x4<S> {
self.simd.unzip_high_u32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_u32x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u32x4(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -1968,6 +2024,14 @@ impl<S: Simd> crate::SimdFloat<S> for f64x2<S> {
self.simd.unzip_high_f64x2(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_f64x2(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_f64x2(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn max(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.max_f64x2(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -2351,6 +2415,14 @@ impl<S: Simd> crate::SimdFloat<S> for f32x8<S> {
self.simd.unzip_high_f32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_f32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_f32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn max(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.max_f32x8(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -2586,6 +2658,14 @@ impl<S: Simd> crate::SimdInt<S> for i8x32<S> {
self.simd.unzip_high_i8x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i8x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_i8x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i8x32(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -2769,6 +2849,14 @@ impl<S: Simd> crate::SimdInt<S> for u8x32<S> {
self.simd.unzip_high_u8x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u8x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_u8x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u8x32(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -3121,6 +3209,15 @@ impl<S: Simd> crate::SimdInt<S> for i16x16<S> {
self.simd.unzip_high_i16x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i16x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd
.deinterleave_i16x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i16x16(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -3309,6 +3406,15 @@ impl<S: Simd> crate::SimdInt<S> for u16x16<S> {
self.simd.unzip_high_u16x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u16x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd
.deinterleave_u16x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u16x16(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -3656,6 +3762,14 @@ impl<S: Simd> crate::SimdInt<S> for i32x8<S> {
self.simd.unzip_high_i32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_i32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i32x8(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -3851,6 +3965,14 @@ impl<S: Simd> crate::SimdInt<S> for u32x8<S> {
self.simd.unzip_high_u32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_u32x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u32x8(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -4222,6 +4344,14 @@ impl<S: Simd> crate::SimdFloat<S> for f64x4<S> {
self.simd.unzip_high_f64x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_f64x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_f64x4(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn max(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.max_f64x4(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -4625,6 +4755,15 @@ impl<S: Simd> crate::SimdFloat<S> for f32x16<S> {
self.simd.unzip_high_f32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_f32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd
.deinterleave_f32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn max(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.max_f32x16(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -4854,6 +4993,14 @@ impl<S: Simd> crate::SimdInt<S> for i8x64<S> {
self.simd.unzip_high_i8x64(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i8x64(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_i8x64(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i8x64(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -5031,6 +5178,14 @@ impl<S: Simd> crate::SimdInt<S> for u8x64<S> {
self.simd.unzip_high_u8x64(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u8x64(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_u8x64(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u8x64(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -5371,6 +5526,15 @@ impl<S: Simd> crate::SimdInt<S> for i16x32<S> {
self.simd.unzip_high_i16x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i16x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd
.deinterleave_i16x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i16x32(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -5553,6 +5717,15 @@ impl<S: Simd> crate::SimdInt<S> for u16x32<S> {
self.simd.unzip_high_u16x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u16x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd
.deinterleave_u16x32(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u16x32(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -5893,6 +6066,15 @@ impl<S: Simd> crate::SimdInt<S> for i32x16<S> {
self.simd.unzip_high_i32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_i32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd
.deinterleave_i32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_i32x16(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -6087,6 +6269,15 @@ impl<S: Simd> crate::SimdInt<S> for u32x16<S> {
self.simd.unzip_high_u32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_u32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd
.deinterleave_u32x16(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn min(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.min_u32x16(self, rhs.simd_into(self.simd))
}
Expand Down Expand Up @@ -6446,6 +6637,14 @@ impl<S: Simd> crate::SimdFloat<S> for f64x8<S> {
self.simd.unzip_high_f64x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn interleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.interleave_f64x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn deinterleave(self, rhs: impl SimdInto<Self, S>) -> (Self, Self) {
self.simd.deinterleave_f64x8(self, rhs.simd_into(self.simd))
}
#[inline(always)]
fn max(self, rhs: impl SimdInto<Self, S>) -> Self {
self.simd.max_f64x8(self, rhs.simd_into(self.simd))
}
Expand Down
Loading
Loading