Wasm32 SIMD implementation

This code is based on rust_sse2.rs of the same distribution, and is
subject to further improvements. Some comments are left intact even if
their applicability is questioned.

SIMD implementation is gated by `wasm32-simd` feature, portable version
is used otherwise.

Performance measurements with a primitive benchmark with ~16Kb of data:

| M1 native     | 11,610 ns |
| M1 WASM SIMD  | 13,355 ns |
| M1 WASM       | 22,037 ns |
| x64 native    |  6,713 ns |
| x64 WASM SIMD | 11,985 ns |
| x64 WASM      | 25,978 ns |

wasmtime v12.0.1 was used on both platforms.

Closes #187.

1 files changed, 19 insertions, 0 deletions

diff --git a/build.rs b/build.rs
index 3f23161..84717f9 100644
--- a/build.rs
+++ b/build.rs
@@ -21,6 +21,10 @@ fn is_no_neon() -> bool {
     defined("CARGO_FEATURE_NO_NEON")
 }
 
+fn is_wasm32_simd() -> bool {
+    defined("CARGO_FEATURE_WASM32_SIMD")
+}
+
 fn is_ci() -> bool {
     defined("BLAKE3_CI")
 }
@@ -91,6 +95,10 @@ fn is_armv7() -> bool {
     target_components()[0] == "armv7"
 }
 
+fn is_wasm32() -> bool {
+    target_components()[0] == "wasm32"
+}
+
 fn endianness() -> String {
     let endianness = env::var("CARGO_CFG_TARGET_ENDIAN").unwrap();
     assert!(endianness == "little" || endianness == "big");
@@ -285,6 +293,13 @@ fn build_neon_c_intrinsics() {
     build.compile("blake3_neon");
 }
 
+fn build_wasm32_simd() {
+    assert!(is_wasm32());
+    // No C code to compile here. Set the cfg flags that enable the WASM SIMD.
+    // The regular Cargo build will compile it.
+    println!("cargo:rustc-cfg=blake3_wasm32_simd");
+}
+
 fn main() -> Result<(), Box<dyn std::error::Error>> {
     // As of Rust 1.80, unrecognized config names are warnings. Give Cargo all of our config names.
     let all_cfgs = [
@@ -341,6 +356,10 @@ fn main() -> Result<(), Box<dyn std::error::Error>> {
         build_neon_c_intrinsics();
     }
 
+    if is_wasm32() && is_wasm32_simd() {
+        build_wasm32_simd();
+    }
+
     // The `cc` crate doesn't automatically emit rerun-if directives for the
     // environment variables it supports, in particular for $CC. We expect to
     // do a lot of benchmarking across different compilers, so we explicitly