Merge pull request #5127 from ethereum/exprUnbreaker

[Yul] Expression joiner.

20 files changed, 545 insertions, 0 deletions

diff --git a/libjulia/optimiser/ExpressionJoiner.cpp b/libjulia/optimiser/ExpressionJoiner.cpp
new file mode 100644
index 00000000..5b92e332
--- /dev/null
+++ b/libjulia/optimiser/ExpressionJoiner.cpp
@@ -0,0 +1,166 @@
+/*
+	This file is part of solidity.
+
+	solidity is free software: you can redistribute it and/or modify
+	it under the terms of the GNU General Public License as published by
+	the Free Software Foundation, either version 3 of the License, or
+	(at your option) any later version.
+
+	solidity is distributed in the hope that it will be useful,
+	but WITHOUT ANY WARRANTY; without even the implied warranty of
+	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+	GNU General Public License for more details.
+
+	You should have received a copy of the GNU General Public License
+	along with solidity.  If not, see <http://www.gnu.org/licenses/>.
+*/
+/**
+ * Optimiser component that undoes what the ExpressionSplitter did, i.e.
+ * it more or less inlines variable declarations.
+ */
+
+#include <libjulia/optimiser/ExpressionJoiner.h>
+
+#include <libjulia/optimiser/NameCollector.h>
+#include <libjulia/optimiser/Utilities.h>
+#include <libjulia/Exceptions.h>
+
+#include <libsolidity/inlineasm/AsmData.h>
+
+#include <libdevcore/CommonData.h>
+
+#include <boost/range/adaptor/reversed.hpp>
+
+using namespace std;
+using namespace dev;
+using namespace dev::julia;
+using namespace dev::solidity;
+
+void ExpressionJoiner::operator()(FunctionalInstruction& _instruction)
+{
+	handleArguments(_instruction.arguments);
+}
+
+void ExpressionJoiner::operator()(FunctionCall& _funCall)
+{
+	handleArguments(_funCall.arguments);
+}
+
+void ExpressionJoiner::operator()(If& _if)
+{
+	visit(*_if.condition);
+	(*this)(_if.body);
+}
+
+void ExpressionJoiner::operator()(Switch& _switch)
+{
+	visit(*_switch.expression);
+	for (auto& _case: _switch.cases)
+		// Do not visit the case expression, nothing to unbreak there.
+		(*this)(_case.body);
+}
+
+void ExpressionJoiner::operator()(Block& _block)
+{
+	resetLatestStatementPointer();
+	for (size_t i = 0; i < _block.statements.size(); ++i)
+	{
+		visit(_block.statements[i]);
+		m_currentBlock = &_block;
+		m_latestStatementInBlock = i;
+	}
+
+	removeEmptyBlocks(_block);
+	resetLatestStatementPointer();
+}
+
+void ExpressionJoiner::visit(Expression& _e)
+{
+	if (_e.type() == typeid(Identifier))
+	{
+		Identifier const& identifier = boost::get<Identifier>(_e);
+		if (isLatestStatementVarDeclOf(identifier) && m_references[identifier.name] == 1)
+		{
+			VariableDeclaration& varDecl = boost::get<VariableDeclaration>(*latestStatement());
+			assertThrow(varDecl.variables.size() == 1, OptimizerException, "");
+			assertThrow(varDecl.value, OptimizerException, "");
+
+			_e = std::move(*varDecl.value);
+			// Delete the variable declaration (also get the moved-from structure back into a sane state)
+			*latestStatement() = Block();
+
+			decrementLatestStatementPointer();
+		}
+	}
+	else
+		ASTModifier::visit(_e);
+}
+
+void ExpressionJoiner::run(Block& _ast)
+{
+	ExpressionJoiner{_ast}(_ast);
+}
+
+ExpressionJoiner::ExpressionJoiner(Block& _ast)
+{
+	ReferencesCounter counter;
+	counter(_ast);
+	m_references = counter.references();
+}
+
+void ExpressionJoiner::handleArguments(vector<Expression>& _arguments)
+{
+	// We have to fill from left to right, but we can only
+	// fill if everything to the right is just an identifier
+	// or a literal.
+	// Also we only descend into function calls if everything
+	// on the right is an identifier or literal.
+
+	size_t i = _arguments.size();
+	for (Expression const& arg: _arguments | boost::adaptors::reversed)
+	{
+		--i;
+		if (arg.type() != typeid(Identifier) && arg.type() != typeid(Literal))
+			break;
+	}
+	// i points to the last element that is neither an identifier nor a literal,
+	// or to the first element if all of them are identifiers or literals.
+
+	for (; i < _arguments.size(); ++i)
+		visit(_arguments.at(i));
+}
+
+void ExpressionJoiner::decrementLatestStatementPointer()
+{
+	if (!m_currentBlock)
+		return;
+	if (m_latestStatementInBlock > 0)
+		--m_latestStatementInBlock;
+	else
+		resetLatestStatementPointer();
+}
+
+void ExpressionJoiner::resetLatestStatementPointer()
+{
+	m_currentBlock = nullptr;
+	m_latestStatementInBlock = size_t(-1);
+}
+
+Statement* ExpressionJoiner::latestStatement()
+{
+	if (!m_currentBlock)
+		return nullptr;
+	else
+		return &m_currentBlock->statements.at(m_latestStatementInBlock);
+}
+
+bool ExpressionJoiner::isLatestStatementVarDeclOf(Identifier const& _identifier)
+{
+	Statement const* statement = latestStatement();
+	if (!statement || statement->type() != typeid(VariableDeclaration))
+		return false;
+	VariableDeclaration const& varDecl = boost::get<VariableDeclaration>(*statement);
+	if (varDecl.variables.size() != 1 || !varDecl.value)
+		return false;
+	return varDecl.variables.at(0).name == _identifier.name;
+}
diff --git a/libjulia/optimiser/ExpressionJoiner.h b/libjulia/optimiser/ExpressionJoiner.h
new file mode 100644
index 00000000..91b45b49
--- /dev/null
+++ b/libjulia/optimiser/ExpressionJoiner.h
@@ -0,0 +1,102 @@
+/*
+	This file is part of solidity.
+
+	solidity is free software: you can redistribute it and/or modify
+	it under the terms of the GNU General Public License as published by
+	the Free Software Foundation, either version 3 of the License, or
+	(at your option) any later version.
+
+	solidity is distributed in the hope that it will be useful,
+	but WITHOUT ANY WARRANTY; without even the implied warranty of
+	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+	GNU General Public License for more details.
+
+	You should have received a copy of the GNU General Public License
+	along with solidity.  If not, see <http://www.gnu.org/licenses/>.
+*/
+/**
+ * Optimiser component that undoes what the ExpressionSplitter did, i.e.
+ * it more or less inlines variable declarations.
+ */
+#pragma once
+
+#include <libjulia/ASTDataForward.h>
+
+#include <libjulia/optimiser/ASTWalker.h>
+
+#include <map>
+
+namespace dev
+{
+namespace julia
+{
+
+class NameCollector;
+
+
+/**
+ * Optimiser component that modifies an AST in place, turning sequences
+ * of variable declarations into complex expressions, if the variables
+ * are declared in the right order. This component does the opposite
+ * of ExpressionSplitter.
+ * Since the order of opcode or function evaluation is unchanged,
+ * this transformation does not need to care about conflicting opcodes.
+ *
+ * Code of the form
+ *
+ * let a1 := mload(y)
+ * let a2 := mul(x, 4)
+ * sstore(a2, a1)
+ *
+ * is transformed into
+ *
+ * sstore(mul(x, 4), mload(y))
+ *
+ * The transformation is not applied to loop conditions, because those are
+ * evaluated with each loop.
+ *
+ * The component can be applied to sub-blocks of the AST, you do not
+ * need to pass a full AST.
+ *
+ * Prerequisites: Disambiguator
+ *
+ * Implementation note: We visit the AST, modifying it in place.
+ * The class starts counting references and will only replace variables
+ * that have exactly one reference. It keeps a "latest statement pointer"
+ * which always points to the statement right before the current statement.
+ * Any function call or opcode will reset this pointer. If an identifier
+ * is encountered that was declared in the "latest statement", it is replaced
+ * by the value of the declaration, the "latest statement" is replaced
+ * by an empty block and the pointer is decremented.
+ * A block also resets the latest statement pointer.
+ */
+class ExpressionJoiner: public ASTModifier
+{
+public:
+	virtual void operator()(FunctionalInstruction&) override;
+	virtual void operator()(FunctionCall&) override;
+	virtual void operator()(If&) override;
+	virtual void operator()(Switch&) override;
+	virtual void operator()(Block& _block) override;
+
+	using ASTModifier::visit;
+	virtual void visit(Expression& _e) override;
+
+	static void run(Block& _ast);
+private:
+	explicit ExpressionJoiner(Block& _ast);
+
+	void handleArguments(std::vector<Expression>& _arguments);
+
+	void decrementLatestStatementPointer();
+	void resetLatestStatementPointer();
+	Statement* latestStatement();
+	bool isLatestStatementVarDeclOf(Identifier const& _identifier);
+
+	Block* m_currentBlock = nullptr;
+	size_t m_latestStatementInBlock = 0;
+	std::map<std::string, size_t> m_references;
+};
+
+}
+}
diff --git a/libjulia/optimiser/README.md b/libjulia/optimiser/README.md
index 877f8255..4a50d5ca 100644
--- a/libjulia/optimiser/README.md
+++ b/libjulia/optimiser/README.md
@@ -56,6 +56,47 @@ depend on any side-effects.
 
 As an example, neither ``mload`` nor ``mstore`` would be allowed.
 
+## Expression Splitter
+
+The expression splitter turns expressions like ``add(mload(x), mul(mload(y), 0x20))``
+into a sequence of declarations of unique variables that are assigned sub-expressions
+of that expression so that each function call has only variables or literals
+as arguments.
+
+The above would be transformed into
+
+    {
+        let _1 := mload(y)
+        let _2 := mul(_1, 0x20)
+        let _3 := mload(x)
+        let z := add(_3, _2)
+    }
+
+Note that this transformation does not change the order of opcodes or function calls.
+
+It is not applied to loop conditions, because the loop control flow does not allow
+this "outlining" of the inner expressions in all cases.
+
+The final program should be in a form such that with the exception of loop conditions,
+function calls can only appear in the right-hand side of a variable declaration,
+assignments or expression statements and all arguments have to be constants or variables.
+
+The benefits of this form are that it is much easier to re-order the sequence of opcodes
+and it is also easier to perform function call inlining. The drawback is that
+such code is much harder to read for humans.
+
+## Expression Joiner
+
+This is the opposite operation of the expression splitter. It turns a sequence of
+variable declarations that have exactly one reference into a complex expression.
+This stage again fully preserves the order of function calls and opcode executions.
+It does not make use of any information concerning the commutability of opcodes;
+if moving the value of a variable to its place of use would change the order
+of any function call or opcode execution, the transformation is not performed.
+
+Note that the component will not move the assigned value of a variable assignment
+or a variable that is referenced more than once.
+
 ## Full Function Inliner
 
 ## Rematerialisation
diff --git a/test/libjulia/YulOptimizerTest.cpp b/test/libjulia/YulOptimizerTest.cpp
index 96e36a5b..38765774 100644
--- a/test/libjulia/YulOptimizerTest.cpp
+++ b/test/libjulia/YulOptimizerTest.cpp
@@ -33,6 +33,7 @@
 #include <libjulia/optimiser/Rematerialiser.h>
 #include <libjulia/optimiser/ExpressionSimplifier.h>
 #include <libjulia/optimiser/UnusedPruner.h>
+#include <libjulia/optimiser/ExpressionJoiner.h>
 
 #include <libsolidity/parsing/Scanner.h>
 #include <libsolidity/inlineasm/AsmPrinter.h>
@@ -146,6 +147,11 @@ bool YulOptimizerTest::run(ostream& _stream, string const& _linePrefix, bool con
 		disambiguate();
 		UnusedPruner::runUntilStabilised(*m_ast);
 	}
+	else if (m_optimizerStep == "expressionJoiner")
+	{
+		disambiguate();
+		ExpressionJoiner::run(*m_ast);\
+	}
 	else
 	{
 		FormattedScope(_stream, _formatted, {formatting::BOLD, formatting::RED}) << _linePrefix << "Invalid optimizer step: " << m_optimizerStep << endl;
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/if_condition.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/if_condition.yul
new file mode 100644
index 00000000..a1349511
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/if_condition.yul
@@ -0,0 +1,21 @@
+{
+	let a := mload(3)
+	let b := sload(a)
+	let c := mload(7)
+	let d := add(c, b)
+	if d {
+		let x := mload(3)
+		let y := add(x, 3)
+	}
+	let z := 3
+	let t := add(z, 9)
+}
+// ----
+// expressionJoiner
+// {
+//     if add(mload(7), sload(mload(3)))
+//     {
+//         let y := add(mload(3), 3)
+//     }
+//     let t := add(3, 9)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/muli_wrong_order3.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/muli_wrong_order3.yul
new file mode 100644
index 00000000..0d407c7c
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/muli_wrong_order3.yul
@@ -0,0 +1,13 @@
+{
+	let a := mload(3)
+	let b := mload(6)
+	let x := mul(add(b, a), mload(2))
+	sstore(x, 3)
+}
+// ----
+// expressionJoiner
+// {
+//     let a := mload(3)
+//     let b := mload(6)
+//     sstore(mul(add(b, a), mload(2)), 3)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/multi.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/multi.yul
new file mode 100644
index 00000000..fd53ca51
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/multi.yul
@@ -0,0 +1,11 @@
+{
+	let a := mload(2)
+	let b := mload(6)
+	let x := mul(add(b, a), 2)
+	sstore(x, 3)
+}
+// ----
+// expressionJoiner
+// {
+//     sstore(mul(add(mload(6), mload(2)), 2), 3)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/multi_reference.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/multi_reference.yul
new file mode 100644
index 00000000..078a12a5
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/multi_reference.yul
@@ -0,0 +1,11 @@
+{
+	// This is not joined because a is referenced multiple times
+	let a := mload(2)
+	let b := add(a, a)
+}
+// ----
+// expressionJoiner
+// {
+//     let a := mload(2)
+//     let b := add(a, a)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/multi_wrong_order.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/multi_wrong_order.yul
new file mode 100644
index 00000000..965e07e9
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/multi_wrong_order.yul
@@ -0,0 +1,15 @@
+{
+	// We have an interleaved "add" here, so we cannot inline "a"
+	// (note that this component does not analyze whether
+	// functions are pure or not)
+	let a := mload(2)
+	let b := mload(6)
+	let x := mul(a, add(2, b))
+	sstore(x, 3)
+}
+// ----
+// expressionJoiner
+// {
+//     let a := mload(2)
+//     sstore(mul(a, add(2, mload(6))), 3)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/multi_wrong_order2.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/multi_wrong_order2.yul
new file mode 100644
index 00000000..c577e182
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/multi_wrong_order2.yul
@@ -0,0 +1,12 @@
+{
+	let a := mload(2)
+	let b := mload(6)
+	let x := mul(add(a, b), 2)
+	sstore(x, 3)
+}
+// ----
+// expressionJoiner
+// {
+//     let a := mload(2)
+//     sstore(mul(add(a, mload(6)), 2), 3)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_across_blocks.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_across_blocks.yul
new file mode 100644
index 00000000..a781bb2a
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_across_blocks.yul
@@ -0,0 +1,19 @@
+{
+	// The component will remove the empty block after
+	// it has handled the outer block.
+	// The idea behind this test is that the component
+	// does not perform replacements across blocks because
+	// they usually have contents, but adding contents
+	// will reduce the scope of the test.
+	let a := mload(2)
+	let x := calldataload(a)
+	{
+	}
+	sstore(x, 3)
+}
+// ----
+// expressionJoiner
+// {
+//     let x := calldataload(mload(2))
+//     sstore(x, 3)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_in_loop_condition1.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_in_loop_condition1.yul
new file mode 100644
index 00000000..75218a5c
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_in_loop_condition1.yul
@@ -0,0 +1,15 @@
+{
+	for { let b := mload(1) } b {} {}
+}
+// ----
+// expressionJoiner
+// {
+//     for {
+//         let b := mload(1)
+//     }
+//     b
+//     {
+//     }
+//     {
+//     }
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_in_loop_condition2.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_in_loop_condition2.yul
new file mode 100644
index 00000000..d5f7d8fa
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/no_replacement_in_loop_condition2.yul
@@ -0,0 +1,16 @@
+{
+	let a := mload(0)
+	for { } a {} {}
+}
+// ----
+// expressionJoiner
+// {
+//     let a := mload(0)
+//     for {
+//     }
+//     a
+//     {
+//     }
+//     {
+//     }
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/only_assignment.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/only_assignment.yul
new file mode 100644
index 00000000..c7411211
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/only_assignment.yul
@@ -0,0 +1,16 @@
+{
+	// This is not joined because a is referenced multiple times
+	function f(a) -> x {
+		a := mload(2)
+		x := add(a, 3)
+	}
+}
+// ----
+// expressionJoiner
+// {
+//     function f(a) -> x
+//     {
+//         a := mload(2)
+//         x := add(a, 3)
+//     }
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/reassignment.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/reassignment.yul
new file mode 100644
index 00000000..1e502353
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/reassignment.yul
@@ -0,0 +1,13 @@
+{
+	// This is not joined because a is referenced multiple times
+	let a := mload(2)
+	let b := mload(a)
+	a := 4
+}
+// ----
+// expressionJoiner
+// {
+//     let a := mload(2)
+//     let b := mload(a)
+//     a := 4
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/simple.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/simple.yul
new file mode 100644
index 00000000..b03bcf45
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/simple.yul
@@ -0,0 +1,10 @@
+{
+	let a := mload(2)
+	let x := calldataload(a)
+	sstore(x, 3)
+}
+// ----
+// expressionJoiner
+// {
+//     sstore(calldataload(mload(2)), 3)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/single_wrong_order.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/single_wrong_order.yul
new file mode 100644
index 00000000..3b433f78
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/single_wrong_order.yul
@@ -0,0 +1,13 @@
+{
+	let a := mload(3)
+	let b := sload(a)
+	let c := mload(7)
+	let d := add(b, c)
+	sstore(d, 0)
+}
+// ----
+// expressionJoiner
+// {
+//     let b := sload(mload(3))
+//     sstore(add(b, mload(7)), 0)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/smoke.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/smoke.yul
new file mode 100644
index 00000000..c0e2c6f2
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/smoke.yul
@@ -0,0 +1,5 @@
+{ }
+// ----
+// expressionJoiner
+// {
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/switch_expression.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/switch_expression.yul
new file mode 100644
index 00000000..0e4e540e
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/switch_expression.yul
@@ -0,0 +1,28 @@
+{
+	let a := mload(3)
+	let b := sload(a)
+	let c := mload(7)
+	let d := add(c, b)
+	switch d
+	case 3 {
+		let x := mload(3)
+		let y := add(x, 3)
+	}
+	default {
+		sstore(1, 0)
+	}
+	let z := 3
+	let t := add(z, 9)
+}
+// ----
+// expressionJoiner
+// {
+//     switch add(mload(7), sload(mload(3)))
+//     case 3 {
+//         let y := add(mload(3), 3)
+//     }
+//     default {
+//         sstore(1, 0)
+//     }
+//     let t := add(3, 9)
+// }
diff --git a/test/libjulia/yulOptimizerTests/expressionJoiner/triple.yul b/test/libjulia/yulOptimizerTests/expressionJoiner/triple.yul
new file mode 100644
index 00000000..7b722be1
--- /dev/null
+++ b/test/libjulia/yulOptimizerTests/expressionJoiner/triple.yul
@@ -0,0 +1,12 @@
+{
+	let a := mload(2)
+	let b := mload(6)
+	let c := mload(7)
+	let x := mul(add(c, b), a)
+	sstore(x, 3)
+}
+// ----
+// expressionJoiner
+// {
+//     sstore(mul(add(mload(7), mload(6)), mload(2)), 3)
+// }