Add AVX-512 support

.github/workflows/ci.yml

@@ -3,12 +3,12 @@ env:
   # version like 1.70. Note that we only specify MAJOR.MINOR and not PATCH so that bugfixes still
   # come automatically. If the version specified here is no longer the latest stable version,
   # then please feel free to submit a PR that adjusts it along with the potential clippy fixes.
-  RUST_STABLE_VER: "1.88" # In quotes because otherwise (e.g.) 1.70 would be interpreted as 1.7
+  RUST_STABLE_VER: "1.89" # In quotes because otherwise (e.g.) 1.70 would be interpreted as 1.7
   # The purpose of checking with the minimum supported Rust toolchain is to detect its staleness.
   # If the compilation fails, then the version specified here needs to be bumped up to reality.
   # Be sure to also update the rust-version property in the workspace Cargo.toml file,
   # plus all the README.md files of the affected packages.
-  RUST_MIN_VER: "1.88"
+  RUST_MIN_VER: "1.89"
   # List of packages that will be checked with the minimum supported Rust version.
   # This should be limited to packages that are intended for publishing.
   RUST_MIN_VER_PKGS: "-p fearless_simd"
@@ -275,8 +275,7 @@ jobs:
       - name: run tests on CPU with AVX-512
         # Github Actions doesn't give us AVX-512 so this is the only way to exercise AVX-512 codepaths on CI.
         # -icl stands for Ice Lake. Technically Skylake added AVX-512 first, but it's mostly useless there due to
-        # downclocking. When we do eventually add explicit AVX-512 support, we'll likely target the Ice Lake feature
-        # level.
+        # downclocking, so our explicit AVX-512 level targets Ice Lake.
         run: ${SDE_PKG}-lin/sde64 -icl -- cargo test $CARGO_TEST_ARGS
 
   test-aarch64-qemu:

CHANGELOG.md

@@ -11,7 +11,15 @@ You can find its changes [documented below](#050-2026-06-18).
 
 ## [Unreleased]
 
-This release has an [MSRV][] of 1.88.
+This release has an [MSRV][] of 1.89.
+
+### Added
+
+- Added Ice Lake-class AVX-512 support with a generated `Avx512` level and 512-bit native-width vector types.
+
+### Changed
+
+- The MSRV is now Rust 1.89.
 
 ## [0.5.0][] (2026-06-18)
 

Cargo.toml

@@ -13,7 +13,7 @@ license = "Apache-2.0 OR MIT"
 repository = "https://github.com/linebender/fearless_simd"
 # Keep in sync with RUST_MIN_VER in .github/workflows/ci.yml, with the relevant README.md files
 # and with the MSRV in the `Unreleased` section of CHANGELOG.md.
-rust-version = "1.88"
+rust-version = "1.89"
 
 [workspace.lints]
 
@@ -44,10 +44,11 @@ clippy.collection_is_never_read = "warn"
 clippy.default_trait_access = "warn"
 clippy.dbg_macro = "warn"
 clippy.debug_assert_with_mut_call = "warn"
+clippy.disallowed_methods = "deny"
 clippy.doc_markdown = "warn"
 clippy.fn_to_numeric_cast_any = "warn"
 clippy.infinite_loop = "warn"
-clippy.large_stack_arrays = "warn"
+clippy.large_stack_arrays = "allow"             # appears to be buggy as of 1.93, fixed in 1.95. TODO: re-enable
 clippy.mismatching_type_param_order = "warn"
 clippy.missing_assert_message = "warn"
 clippy.missing_fields_in_debug = "warn"

README.md

@@ -54,7 +54,7 @@ It benefited from conversations with Luca Versari, though he is not responsible
 
 ## Minimum supported Rust Version (MSRV)
 
-This version of Fearless SIMD has been verified to compile with **Rust 1.88** and later.
+This version of Fearless SIMD has been verified to compile with **Rust 1.89** and later.
 
 Future versions of Fearless SIMD might increase the Rust version requirement.
 It will not be treated as a breaking change and as such can even happen with small patch releases.

check_targets.sh

@@ -15,6 +15,8 @@ cargo check -p fearless_simd --target aarch64-linux-android  --features force_su
 cargo check -p fearless_simd --target aarch64-linux-android
 
 # x86_64, at all supported static SIMD levels.
+RUSTFLAGS=-Ctarget-cpu=icelake-server cargo check -p fearless_simd --target x86_64-unknown-linux-gnu
+RUSTFLAGS=-Ctarget-cpu=icelake-server cargo check -p fearless_simd --target x86_64-unknown-linux-gnu --features force_support_fallback
 RUSTFLAGS=-Ctarget-feature=+avx2,+fma cargo check -p fearless_simd --target x86_64-unknown-linux-gnu
 RUSTFLAGS=-Ctarget-feature=+avx2,+fma cargo check -p fearless_simd --target x86_64-unknown-linux-gnu --features force_support_fallback
 RUSTFLAGS=-Ctarget-feature=+sse4.2 cargo check -p fearless_simd --target x86_64-unknown-linux-gnu

fearless_simd/README.md

@@ -146,7 +146,7 @@ case. There's also Q&A on [Zulip](https://xi.zulipchat.com/#narrow/channel/51423
 
 ## Instruction set support
 
-- x86/x86-64: [v2](https://en.wikipedia.org/wiki/X86-64#Microarchitecture_levels) (SSE4.2), [v3](https://en.wikipedia.org/wiki/X86-64#Microarchitecture_levels) (AVX2)
+- x86/x86-64: [v2](https://en.wikipedia.org/wiki/X86-64#Microarchitecture_levels) (SSE4.2), [v3](https://en.wikipedia.org/wiki/X86-64#Microarchitecture_levels) (AVX2), [Ice Lake](https://en.wikipedia.org/wiki/AVX-512#CPUs_with_AVX-512) (AVX-512, avoiding early slow implementations)
 - Aarch64: Baseline [NEON](https://en.wikipedia.org/wiki/Arm_architecture_family#Advanced_SIMD_(Neon))
 - WebAssembly: [128-bit packed SIMD](https://github.com/WebAssembly/spec/blob/main/proposals/simd/SIMD.md), [relaxed SIMD](https://github.com/WebAssembly/relaxed-simd/blob/main/proposals/relaxed-simd/Overview.md)
 
@@ -204,7 +204,7 @@ At least one of `std` and `libm` is required; `std` overrides `libm`.
 
 ## Minimum supported Rust Version (MSRV)
 
-This version of Fearless SIMD has been verified to compile with **Rust 1.88** and later.
+This version of Fearless SIMD has been verified to compile with **Rust 1.89** and later.
 
 Future versions of Fearless SIMD might increase the Rust version requirement.
 It will not be treated as a breaking change and as such can even happen with small patch releases.

fearless_simd/src/generated.rs

@@ -41,6 +41,8 @@
 
 #[cfg(any(target_arch = "x86", target_arch = "x86_64"))]
 mod avx2;
+#[cfg(any(target_arch = "x86", target_arch = "x86_64"))]
+mod avx512;
 mod fallback;
 #[cfg(target_arch = "aarch64")]
 mod neon;
@@ -54,6 +56,8 @@ mod wasm;
 
 #[cfg(any(target_arch = "x86", target_arch = "x86_64"))]
 pub use avx2::*;
+#[cfg(any(target_arch = "x86", target_arch = "x86_64"))]
+pub use avx512::*;
 pub use fallback::*;
 #[cfg(target_arch = "aarch64")]
 pub use neon::*;