[perf] WASM hashing function

packages/hash-rust/.gitignore

@@ -0,0 +1,6 @@
+/target
+**/*.rs.bk
+Cargo.lock
+bin/
+pkg/
+wasm-pack.log

packages/hash-rust/Cargo.toml

@@ -0,0 +1,30 @@
+[package]
+name = "murmur"
+version = "0.1.0"
+authors = ["Rom Grk <[email protected]>"]
+edition = "2018"
+
+[lib]
+crate-type = ["cdylib", "rlib"]
+
+[dependencies]
+cityhasher = "0.1.0"
+wasm-bindgen = "0.2.84"
+
+[dependencies.xxhash-rust]
+version = "0.8.12"
+features = ["xxh32", "xxh3"]
+
+[dev-dependencies]
+wasm-bindgen-test = "0.3.34"
+
+[package.metadata.wasm-pack.profile.dev]
+wasm-opt = ['--enable-simd']
+[package.metadata.wasm-pack.profile.profiling]
+wasm-opt = ['--enable-simd']
+[package.metadata.wasm-pack.profile.release]
+wasm-opt = ['--enable-simd']
+
+[profile.release]
+# Tell `rustc` to optimize for small code size.
+opt-level = "s"

packages/hash-rust/README.md

@@ -0,0 +1,5 @@
+# hash-rust
+
+Bootstraped from a standard `wasm-pack` template.
+
+Build with `wasm-pack build --target nodejs`

packages/hash-rust/index.js

@@ -0,0 +1,27 @@
+let imports = {}
+let wasm
+
+const path = require('path').join(__dirname, './pkg/murmur_bg.wasm')
+const bytes = require('fs').readFileSync(path)
+const wasmModule = new WebAssembly.Module(bytes)
+const wasmInstance = new WebAssembly.Instance(wasmModule, imports)
+wasm = wasmInstance.exports
+
+// Manually written to encode faster than wasm_bindgen does.
+
+let MEM_START = 32 // Alignement for the bytes slice
+
+let memory = new Uint8Array(wasm.memory.buffer).subarray(MEM_START)
+let cachedTextEncoder = new TextEncoder('utf-8')
+
+module.exports.xxh = wrapStringToU32('xxh')
+module.exports.murmur2 = wrapStringToU32('murmur2')
+
+function wrapStringToU32(name) {
+  return input => {
+    const ptr0 = MEM_START
+    const len0 = cachedTextEncoder.encodeInto(input, memory).written
+    const ret = wasm[name](ptr0, len0)
+    return ret
+  }
+}

packages/hash-rust/src/lib.rs

@@ -0,0 +1,63 @@
+use wasm_bindgen::prelude::*;
+use xxhash_rust::xxh32::xxh32;
+use xxhash_rust::xxh3::xxh3_64;
+
+// xxh via xxhash_rust
+
+#[wasm_bindgen]
+pub fn xxh(bytes: *const u8, length: usize) -> u32 {
+  let data = unsafe { std::slice::from_raw_parts(
+    std::mem::transmute::<_, *const u8>(bytes),
+    length,
+  ) };
+
+  return xxh32(data, 0);
+}
+
+
+// murmur2 implementation based on emotion's implementation (differs from original)
+
+const M: u32 = 0x5bd1e995;
+const R: u32 = 24;
+
+#[wasm_bindgen]
+pub fn murmur2(bytes: *const u8, mut length: usize) -> u32 {
+  let data = unsafe { std::slice::from_raw_parts(
+    std::mem::transmute::<*const u8, *const u32>(bytes),
+    length / 4,
+  ) };
+
+  let mut h = 0; // Not initialized to match emotion's implementation
+  let mut i = 0;
+
+  while length >= 4 {
+    let mut k = unsafe { *data.get_unchecked(i) };
+
+    k *= M;
+    k ^= k >> R;
+    k *= M;
+
+    h *= M;
+    h ^= k;
+
+    i += 1;
+    length -= 4;
+  }
+
+  if length >= 3 {
+    h ^= (unsafe { *bytes.add(i + 2) as u32 } & 0xff) << 16;
+  }
+  if length >= 2 {
+    h ^= (unsafe { *bytes.add(i + 1) as u32 } & 0xff) << 8;
+  }
+  if length >= 1 {
+    h ^= (unsafe { *bytes.add(i + 0) as u32 } & 0xff);
+    h *= M;
+  }
+
+  h ^= h >> 13;
+  h *= M;
+  h = h ^ (h >> 15);
+
+  return h;
+}