Improve Runtime execution by caching runtime lookup (#276)

* Improve Runtime execution by caching runtime lookup

Cache whether the native or wasm runtime should be used for a given
code and if the latter, keep around the parsed wasmi::Module for faster
execution.

* Additional comment and code style fixes

* Fallback to WASM runtime if we can't call the version function

* The previous assumption that the wasm-code was compiled with rustc might be wrong now, that the code comes from the blockchain. Added Notes about that.

substrate/Cargo.lock

@@ -2165,7 +2165,9 @@ dependencies = [
  "ed25519 0.1.0",
  "error-chain 0.12.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "hex-literal 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
+ "lazy_static 1.0.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "log 0.3.9 (registry+https://github.com/rust-lang/crates.io-index)",
+ "parking_lot 0.4.8 (registry+https://github.com/rust-lang/crates.io-index)",
  "rustc-hex 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "serde 1.0.64 (registry+https://github.com/rust-lang/crates.io-index)",
  "serde_derive 1.0.64 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -2176,6 +2178,7 @@ dependencies = [
  "substrate-serializer 0.1.0",
  "substrate-state-machine 0.1.0",
  "triehash 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
+ "twox-hash 1.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "wabt 0.4.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "wasmi 0.3.0 (registry+https://github.com/rust-lang/crates.io-index)",
 ]

substrate/substrate/executor/Cargo.toml

@@ -19,6 +19,9 @@ byteorder = "1.1"
 rustc-hex = "1.0.0"
 triehash = "0.1.0"
 hex-literal = "0.1.0"
+twox-hash = "1.1.0"
+lazy_static = "1.0"
+parking_lot = "*"
 log = "0.3"
 
 [dev-dependencies]

substrate/substrate/executor/src/lib.rs

@@ -41,8 +41,13 @@ extern crate wasmi;
 extern crate byteorder;
 extern crate rustc_hex;
 extern crate triehash;
+extern crate parking_lot;
+extern crate twox_hash;
 #[macro_use] extern crate log;
 
+#[macro_use]
+extern crate lazy_static;
+
 #[macro_use]
 extern crate error_chain;
 

substrate/substrate/executor/src/native_executor.rs

@@ -17,10 +17,35 @@
 use error::{Error, ErrorKind, Result};
 use state_machine::{CodeExecutor, Externalities};
 use wasm_executor::WasmExecutor;
+use wasmi::Module as WasmModule;
 use runtime_version::RuntimeVersion;
+use std::collections::HashMap;
 use codec::Slicable;
+use twox_hash::XxHash;
+use std::hash::Hasher;
+use parking_lot::Mutex;
 use RuntimeInfo;
 
+// For the internal Runtime Cache
+enum RunWith {
+	NativeRuntime,
+	WasmRuntime(WasmModule)
+}
+
+lazy_static! {
+	static ref RUNTIMES_CACHE: Mutex<HashMap<u64, RunWith>> = Mutex::new(HashMap::new());
+}
+
+// helper function to generate low-over-head caching_keys
+// it is asserted that part of the audit process that any potential on-chain code change
+// will have done is to ensure that the two-x hash is different to that of any other
+// :code value from the same chain
+fn gen_cache_key(code: &[u8]) -> u64 {
+	let mut h = XxHash::with_seed(0);
+	h.write(code);
+	h.finish()
+}
+
 fn safe_call<F, U>(f: F) -> Result<U>
 	where F: ::std::panic::UnwindSafe + FnOnce() -> U
 {
@@ -60,6 +85,11 @@ pub struct NativeExecutor<D: NativeExecutionDispatch + Sync + Send> {
 impl<D: NativeExecutionDispatch + Sync + Send> NativeExecutor<D> {
 	/// Create new instance.
 	pub fn new() -> Self {
+		// FIXME: set this entry at compile time
+		RUNTIMES_CACHE.lock().insert(
+			gen_cache_key(D::native_equivalent()),
+			RunWith::NativeRuntime);
+
 		NativeExecutor {
 			_dummy: Default::default(),
 		}
@@ -88,17 +118,24 @@ impl<D: NativeExecutionDispatch + Sync + Send> CodeExecutor for NativeExecutor<D
 		method: &str,
 		data: &[u8],
 	) -> Result<Vec<u8>> {
-		if code == D::native_equivalent() {
-			// call native
-			D::dispatch(ext, method, data)
-		} else {
-			let version = WasmExecutor.call(ext, code, "version", &[])?;
-			let version = RuntimeVersion::decode(&mut version.as_slice());
-			if version.map_or(false, |v| D::VERSION.can_call_with(&v)) {
-				return D::dispatch(ext, method, data)
-			}
-			// call into wasm.
-			WasmExecutor.call(ext, code, method, data)
+		let mut cache = RUNTIMES_CACHE.lock();
+		let r = cache.entry(gen_cache_key(code))
+			.or_insert_with(|| {
+				let wasm_module = WasmModule::from_buffer(code)
+					.expect("all modules compiled with rustc are valid wasm code; qed");
+
+				if WasmExecutor.call_in_wasm_module(ext, &wasm_module, "version", &[]).ok()
+					.and_then(|version| RuntimeVersion::decode(&mut version.as_slice()))
+					.map_or(false, |v| D::VERSION.can_call_with(&v))
+				{
+					RunWith::NativeRuntime
+				} else {
+					RunWith::WasmRuntime(wasm_module)
+				}
+			});
+		match r {
+			RunWith::NativeRuntime => D::dispatch(ext, method, data),
+			RunWith::WasmRuntime(module) => WasmExecutor.call_in_wasm_module(ext, &module, method, data)
 		}
 	}
 }

substrate/substrate/executor/src/wasm_executor.rs

@@ -468,21 +468,19 @@ impl_function_executor!(this: FunctionExecutor<'e, E>,
 #[derive(Debug, Default, Clone)]
 pub struct WasmExecutor;
 
-impl CodeExecutor for WasmExecutor {
-	type Error = Error;
+impl WasmExecutor {
 
-	fn call<E: Externalities>(
+	/// Call a given method in the given wasm-module runtime.
+	pub fn call_in_wasm_module<E: Externalities>(
 		&self,
 		ext: &mut E,
-		code: &[u8],
+		module: &Module,
 		method: &str,
 		data: &[u8],
 	) -> Result<Vec<u8>> {
-		let module = Module::from_buffer(code).expect("all modules compiled with rustc are valid wasm code; qed");
-
 		// start module instantiation. Don't run 'start' function yet.
 		let intermediate_instance = ModuleInstance::new(
-			&module,
+			module,
 			&ImportsBuilder::new()
 				.with_resolver("env", FunctionExecutor::<E>::resolver())
 		)?;
@@ -492,6 +490,8 @@ impl CodeExecutor for WasmExecutor {
 		let memory = intermediate_instance
 			.not_started_instance()
 			.export_by_name("memory")
+			// TODO: with code coming from the blockchain it isn't strictly been compiled with rustc anymore.
+			// these assumptions are probably not true anymore
 			.expect("all modules compiled with rustc should have an export named 'memory'; qed")
 			.as_memory()
 			.expect("in module generated by rustc export named 'memory' should be a memory; qed")
@@ -529,6 +529,21 @@ impl CodeExecutor for WasmExecutor {
 	}
 }
 
+impl CodeExecutor for WasmExecutor {
+	type Error = Error;
+
+	fn call<E: Externalities>(
+		&self,
+		ext: &mut E,
+		code: &[u8],
+		method: &str,
+		data: &[u8],
+	) -> Result<Vec<u8>> {
+		let module = Module::from_buffer(code)?;
+		self.call_in_wasm_module(ext, &module, method, data)
+	}
+}
+
 #[cfg(test)]
 mod tests {
 	use super::*;