Merge pull request #11026 from ethereum/resolve-keccak

Evaluate ``keccak256(a, const)``, when the value at memory location ``a`` is known at compile time and ``const <= 32``

libyul/optimiser/DataFlowAnalyzer.cpp

@@ -27,6 +27,7 @@
 #include <libyul/AST.h>
 #include <libyul/Dialect.h>
 #include <libyul/Exceptions.h>
+#include <libyul/Utilities.h>
 
 #include <libsolutil/CommonData.h>
 #include <libsolutil/cxx20.h>
@@ -388,6 +389,14 @@ bool DataFlowAnalyzer::inScope(YulString _variableName) const
 	return false;
 }
 
+optional<u256> DataFlowAnalyzer::valueOfIdentifier(YulString const& _name)
+{
+	if (m_value.count(_name))
+		if (Literal const* literal = get_if<Literal>(m_value.at(_name).value))
+			return valueOfLiteral(*literal);
+	return nullopt;
+}
+
 std::optional<pair<YulString, YulString>> DataFlowAnalyzer::isSimpleStore(
 	StoreLoadLocation _location,
 	ExpressionStatement const& _statement
@@ -412,4 +421,3 @@ std::optional<YulString> DataFlowAnalyzer::isSimpleLoad(
 				return key->name;
 	return {};
 }
-

libyul/optimiser/DataFlowAnalyzer.h

@@ -29,6 +29,8 @@
 #include <libyul/AST.h> // Needed for m_zero below.
 #include <libyul/SideEffects.h>
 
+#include <libsolutil/Common.h>
+
 #include <map>
 #include <set>
 
@@ -134,6 +136,9 @@ protected:
 	/// Returns true iff the variable is in scope.
 	bool inScope(YulString _variableName) const;
 
+	/// Returns the literal value of the identifier, if it exists.
+	std::optional<u256> valueOfIdentifier(YulString const& _name);
+
 	enum class StoreLoadLocation {
 		Memory = 0,
 		Storage = 1,

libyul/optimiser/LoadResolver.cpp

@@ -23,13 +23,23 @@
 #include <libyul/optimiser/LoadResolver.h>
 
 #include <libyul/backends/evm/EVMDialect.h>
+#include <libyul/backends/evm/EVMMetrics.h>
 #include <libyul/optimiser/Semantics.h>
 #include <libyul/optimiser/CallGraphGenerator.h>
 #include <libyul/SideEffects.h>
 #include <libyul/AST.h>
+#include <libyul/Utilities.h>
+
+#include <libevmasm/Instruction.h>
+#include <libevmasm/GasMeter.h>
+#include <libsolutil/Keccak256.h>
+
+#include <limits>
 
 using namespace std;
 using namespace solidity;
+using namespace solidity::util;
+using namespace solidity::evmasm;
 using namespace solidity::yul;
 
 void LoadResolver::run(OptimiserStepContext& _context, Block& _ast)
@@ -38,7 +48,8 @@ void LoadResolver::run(OptimiserStepContext& _context, Block& _ast)
 	LoadResolver{
 		_context.dialect,
 		SideEffectsPropagator::sideEffects(_context.dialect, CallGraphGenerator::callGraph(_ast)),
-		!containsMSize
+		containsMSize,
+		_context.expectedExecutionsPerDeployment
 	}(_ast);
 }
 
@@ -47,12 +58,17 @@ void LoadResolver::visit(Expression& _e)
 	DataFlowAnalyzer::visit(_e);
 
 	if (FunctionCall const* funCall = std::get_if<FunctionCall>(&_e))
+	{
 		for (auto location: { StoreLoadLocation::Memory, StoreLoadLocation::Storage })
 			if (funCall->functionName.name == m_loadFunctionName[static_cast<unsigned>(location)])
 			{
 				tryResolve(_e, location, funCall->arguments);
 				break;
 			}
+
+		if (!m_containsMSize && funCall->functionName.name == m_dialect.hashFunction({}))
+			tryEvaluateKeccak(_e, funCall->arguments);
+	}
 }
 
 void LoadResolver::tryResolve(
@@ -71,8 +87,64 @@ void LoadResolver::tryResolve(
 			if (inScope(*value))
 				_e = Identifier{locationOf(_e), *value};
 	}
-	else if (m_optimizeMLoad && _location == StoreLoadLocation::Memory)
+	else if (!m_containsMSize && _location == StoreLoadLocation::Memory)
 		if (auto value = util::valueOrNullptr(m_memory, key))
 			if (inScope(*value))
 				_e = Identifier{locationOf(_e), *value};
 }
+
+void LoadResolver::tryEvaluateKeccak(
+	Expression& _e,
+	std::vector<Expression> const& _arguments
+)
+{
+	// The costs are only correct for hashes of 32 bytes or 1 word (when rounded up).
+	GasMeter gasMeter{
+		dynamic_cast<EVMDialect const&>(m_dialect),
+		!m_expectedExecutionsPerDeployment,
+		m_expectedExecutionsPerDeployment ? *m_expectedExecutionsPerDeployment : 1
+	};
+
+	bigint costOfKeccak = gasMeter.costs(_e);
+	bigint costOfLiteral = gasMeter.costs(
+		Literal{
+			{},
+			LiteralKind::Number,
+			// a dummy 256-bit number to represent the Keccak256 hash.
+			YulString{numeric_limits<u256>::max().str()},
+			{}
+		}
+	);
+
+	// We skip if there are no net gas savings.
+	// Note that for default `m_runs = 200`, the values are
+	// `costOfLiteral = 7200` and `costOfKeccak = 9000` for runtime context.
+	// For creation context: `costOfLiteral = 531` and `costOfKeccak = 90`.
+	if (costOfLiteral > costOfKeccak)
+		return;
+
+	yulAssert(_arguments.size() == 2, "");
+	Identifier const* memoryKey = std::get_if<Identifier>(&_arguments.at(0));
+	Identifier const* length = std::get_if<Identifier>(&_arguments.at(1));
+
+	if (!memoryKey || !length)
+		return;
+
+	auto memoryValue = util::valueOrNullptr(m_memory, memoryKey->name);
+	if (memoryValue && inScope(*memoryValue))
+	{
+		optional<u256> memoryContent = valueOfIdentifier(*memoryValue);
+		optional<u256> byteLength = valueOfIdentifier(length->name);
+		if (memoryContent && byteLength && *byteLength <= 32)
+		{
+			bytes contentAsBytes = toBigEndian(*memoryContent);
+			contentAsBytes.resize(static_cast<size_t>(*byteLength));
+			_e = Literal{
+				locationOf(_e),
+				LiteralKind::Number,
+				YulString{u256(keccak256(contentAsBytes)).str()},
+				m_dialect.defaultType
+			};
+		}
+	}
+}

libyul/optimiser/LoadResolver.h

@@ -32,6 +32,9 @@ namespace solidity::yul
  * Optimisation stage that replaces expressions of type ``sload(x)`` and ``mload(x)`` by the value
  * currently stored in storage resp. memory, if known.
  *
+ * Also evaluates simple ``keccak256(a, c)`` when the value at memory location `a` is known and `c`
+ * is a constant `<= 32`.
+ *
  * Works best if the code is in SSA form.
  *
  * Prerequisite: Disambiguator, ForLoopInitRewriter.
@@ -47,10 +50,12 @@ private:
 	LoadResolver(
 		Dialect const& _dialect,
 		std::map<YulString, SideEffects> _functionSideEffects,
-		bool _optimizeMLoad
+		bool _containsMSize,
+		std::optional<size_t> _expectedExecutionsPerDeployment
 	):
 		DataFlowAnalyzer(_dialect, std::move(_functionSideEffects)),
-		m_optimizeMLoad(_optimizeMLoad)
+		m_containsMSize(_containsMSize),
+		m_expectedExecutionsPerDeployment(std::move(_expectedExecutionsPerDeployment))
 	{}
 
 protected:
@@ -63,7 +68,17 @@ protected:
 		std::vector<Expression> const& _arguments
 	);
 
-	bool m_optimizeMLoad = false;
+	/// Evaluates simple ``keccak256(a, c)`` when the value at memory location ``a`` is known and
+	/// `c` is a constant `<= 32`.
+	void tryEvaluateKeccak(
+		Expression& _e,
+		std::vector<Expression> const& _arguments
+	);
+
+	/// If the AST contains `msize`, then we skip resolving `mload` and `keccak256`.
+	bool m_containsMSize = false;
+	/// The --optimize-runs parameter. Value `nullopt` represents creation code.
+	std::optional<size_t> m_expectedExecutionsPerDeployment;
 };
 
 }

libyul/optimiser/OptimiserStep.h

@@ -37,6 +37,8 @@ struct OptimiserStepContext
 	Dialect const& dialect;
 	NameDispenser& dispenser;
 	std::set<YulString> const& reservedIdentifiers;
+	/// The value nullopt represents creation code
+	std::optional<size_t> expectedExecutionsPerDeployment;
 };
 
 

libyul/optimiser/Suite.cpp

@@ -88,6 +88,7 @@ void OptimiserSuite::run(
 	Object& _object,
 	bool _optimizeStackAllocation,
 	string const& _optimisationSequence,
+	optional<size_t> _expectedExecutionsPerDeployment,
 	set<YulString> const& _externallyUsedIdentifiers
 )
 {
@@ -101,7 +102,7 @@ void OptimiserSuite::run(
 	)(*_object.code));
 	Block& ast = *_object.code;
 
-	OptimiserSuite suite(_dialect, reservedIdentifiers, Debug::None, ast);
+	OptimiserSuite suite(_dialect, reservedIdentifiers, Debug::None, ast, _expectedExecutionsPerDeployment);
 
 	// Some steps depend on properties ensured by FunctionHoister, BlockFlattener, FunctionGrouper and
 	// ForLoopInitRewriter. Run them first to be able to run arbitrary sequences safely.

libyul/optimiser/Suite.h

@@ -58,12 +58,14 @@ public:
 		PrintStep,
 		PrintChanges
 	};
+	/// The value nullopt for `_expectedExecutionsPerDeployment` represents creation code.
 	static void run(
 		Dialect const& _dialect,
 		GasMeter const* _meter,
 		Object& _object,
 		bool _optimizeStackAllocation,
 		std::string const& _optimisationSequence,
+		std::optional<size_t> _expectedExecutionsPerDeployment,
 		std::set<YulString> const& _externallyUsedIdentifiers = {}
 	);
 
@@ -88,10 +90,11 @@ private:
 		Dialect const& _dialect,
 		std::set<YulString> const& _externallyUsedIdentifiers,
 		Debug _debug,
-		Block& _ast
+		Block& _ast,
+		std::optional<size_t> expectedExecutionsPerDeployment
 	):
 		m_dispenser{_dialect, _ast, _externallyUsedIdentifiers},
-		m_context{_dialect, m_dispenser, _externallyUsedIdentifiers},
+		m_context{_dialect, m_dispenser, _externallyUsedIdentifiers, expectedExecutionsPerDeployment},
 		m_debug(_debug)
 	{}