k12: new implementation with parallel processing support (#839)

New implementation supports parallel processing of data and consumes
data as soon as possible, which allows to remove the 8 KiB buffer used
previously. It also introduces separate customized types with owned and
borrowed customization strings, while the base type can not be
customized.

Author: newpavlov

Cargo.lock

@@ -35,5 +35,4 @@ opt-level = 2
 
 [patch.crates-io]
 sha1 = { path = "sha1" }
-turbo-shake = { path = "turbo-shake" }
 whirlpool = { path = "whirlpool" }

Cargo.toml

@@ -10,11 +10,11 @@ repository = "https://github.com/RustCrypto/hashes"
 license = "Apache-2.0 OR MIT"
 keywords = ["hash", "digest"]
 categories = ["cryptography", "no-std"]
-description = "Pure Rust implementation of the KangarooTwelve hash function"
+description = "Implementation of the KangarooTwelve family of extendable-output functions"
 
 [dependencies]
 digest = "0.11"
-turbo-shake = { version = "0.1.0", default-features = false }
+keccak = { version = "0.2", features = ["parallel"] }
 
 [dev-dependencies]
 digest = { version = "0.11", features = ["alloc", "dev"] }
@@ -23,7 +23,7 @@ hex-literal = "1"
 [features]
 default = ["alloc"]
 alloc = ["digest/alloc"]
-zeroize = ["digest/zeroize", "turbo-shake/zeroize"]
+zeroize = ["digest/zeroize"]
 
 [package.metadata.docs.rs]
 all-features = true

k12/Cargo.toml

@@ -6,7 +6,7 @@
 ![Rust Version][rustc-image]
 [![Build Status][build-image]][build-link]
 
-Pure Rust implementation of the [KangarooTwelve] eXtendable-Output Function (XOF).
+Pure Rust implementation of the [KangarooTwelve] family of extendable-output functions (XOF).
 
 ## License
 

k12/README.md

@@ -0,0 +1,14 @@
+#![feature(test)]
+extern crate test;
+
+use test::Bencher;
+
+digest::bench_update!(
+    k12::Kt128::default();
+    kt128_1_10 10;
+    kt128_2_100 100;
+    kt128_3_1k 1000;
+    kt128_4_10k 10_000;
+    kt128_5_100k 100_000;
+    kt128_6_1m 1_000_000;
+);

k12/benches/kt128.rs

@@ -0,0 +1,14 @@
+#![feature(test)]
+extern crate test;
+
+use test::Bencher;
+
+digest::bench_update!(
+    k12::Kt256::default();
+    kt256_1_10 10;
+    kt256_2_100 100;
+    kt256_3_1k 1000;
+    kt256_4_10k 10_000;
+    kt256_5_100k 100_000;
+    kt256_6_1m 1_000_000;
+);

k12/benches/kt256.rs

@@ -0,0 +1,14 @@
+/// Number of permutation rounds used by k12
+pub(crate) const ROUNDS: usize = 12;
+/// Chunk size used by k12
+pub(crate) const CHUNK_SIZE: usize = 1 << 13;
+pub(crate) const CHUNK_SIZE_U64: u64 = CHUNK_SIZE as u64;
+
+pub(crate) const SINGLE_NODE_DS: u8 = 0x07;
+pub(crate) const INTERMEDIATE_NODE_DS: u8 = 0x0B;
+pub(crate) const FINAL_NODE_DS: u8 = 0x06;
+
+pub(crate) const S0_DELIM: u64 = 0x03;
+
+/// Padding byte
+pub(crate) const PAD: u8 = 0x80;