Add feature extraction (#1785)

* Add push0 and push tests

* Fix issue with args

* Add missing files

* Add feature extraction

* Rename file
pull/1784/head
Nikhil Parasaram 1 year ago committed by GitHub
parent ca2dc138f8
commit 96c3ecd4b8
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
  1. 234
      mythril/solidity/features.py
  2. 72
      tests/features_test.py
  3. 19
      tests/testdata/input_contracts/SimpleModifier.sol
  4. 398
      tests/testdata/input_contracts/WalletLibrary.sol
  5. 4
      tests/testdata/input_contracts/exceptions.sol
  6. 2
      tests/testdata/input_contracts/kcalls.sol

@ -0,0 +1,234 @@
TRANSFER_METHODS = ["transfer", "send"]
class SolidityFeatureExtractor:
def __init__(self, ast):
self.ast = ast
def extract_features(self):
function_features = {}
function_nodes = self.find_function_nodes(self.ast)
modifier_vars = {}
for modifier_node in self.find_modifier_nodes(self.ast):
modifier_vars[modifier_node["name"]] = self.find_variables_in_require(
modifier_node
)
modifier_vars[modifier_node["name"]].update(
self.find_variables_in_if(modifier_node)
)
print(modifier_vars)
for node in function_nodes:
function_name = self.get_function_name(node)
contains_selfdestruct = self.contains_selfdestruct(node)
contains_call = self.contains_call(node)
contains_delegatecall = self.contains_delegatecall(node)
contains_callcode = self.contains_callcode(node)
contains_staticcall = self.contains_staticcall(node)
all_require_vars = self.find_variables_in_require(node)
ether_vars = self.extract_address_variable(node)
print(ether_vars)
for modifier in node.get("modifiers", []):
all_require_vars.update(modifier_vars[modifier["modifierName"]["name"]])
is_payable = self.is_function_payable(node)
has_isowner_modifier = self.has_isowner_modifier(node)
contains_assert = self.contains_assert(node)
function_features[function_name] = {
"contains_selfdestruct": contains_selfdestruct,
"contains_call": contains_call,
"is_payable": is_payable,
"has_owner_modifier": has_isowner_modifier,
"contains_assert": contains_assert,
"contains_callcode": contains_callcode,
"contains_delegatecall": contains_delegatecall,
"contains_staticcall": contains_staticcall,
"all_require_vars": all_require_vars,
"transfer_vars": ether_vars,
}
return function_features
def find_function_nodes(self, node):
if node["nodeType"] == "FunctionDefinition":
yield node
if "nodes" in node:
for child_node in node["nodes"]:
yield from self.find_function_nodes(child_node)
def find_modifier_nodes(self, node):
if node["nodeType"] == "ModifierDefinition":
yield node
if "nodes" in node:
for child_node in node["nodes"]:
yield from self.find_modifier_nodes(child_node)
def get_function_name(self, node):
return node["name"]
def contains_command(self, node, command):
if isinstance(node, dict):
if command in node.values():
return True
for value in node.values():
if isinstance(value, (dict, list)):
if self.contains_command(value, command):
return True
elif isinstance(node, list):
for item in node:
if self.contains_command(item, command):
return True
return False
def contains_call(self, node):
return self.contains_command(node, "call")
def is_function_payable(self, node):
return node.get("stateMutability") == "payable"
def has_isowner_modifier(self, node):
if "modifiers" in node:
for modifier in node["modifiers"]:
if modifier["modifierName"]["name"].lower() in ("isowner", "onlyowner"):
return True
return False
def contains_assert(self, node):
return self.contains_command(node, "assert")
def contains_selfdestruct(self, node):
return self.contains_command(node, "selfdestruct")
def contains_delegatecall(self, node):
return self.contains_command(node, "delegatecall")
def contains_callcode(self, node):
return self.contains_command(node, "callcode")
def contains_staticcall(self, node):
return self.contains_command(node, "staticcall")
def contains_require(self, node):
return self.contains_command(node, "require")
def extract_nodes(self, node, command, parent=None):
node_list = []
if isinstance(node, dict):
if command in node.values():
node_list.append((parent, node))
for key, value in node.items():
if isinstance(value, (dict, list)):
node_list.extend(self.extract_nodes(value, command, parent=node))
elif isinstance(node, list):
for item in node:
node_list.extend(self.extract_nodes(item, command, parent=node))
return node_list
def find_all_variables(self, node):
variables = set()
def traverse(node):
if isinstance(node, dict):
for key, value in node.items():
if key == "nodeType" and value == "Identifier":
if "name" in node:
variables.add(node["name"])
elif isinstance(value, (dict, list)):
traverse(value)
elif isinstance(node, list):
for item in node:
traverse(item)
traverse(node)
return variables
def find_variables_in_require(self, node):
nodes = self.extract_nodes(node, "require")
variables = set()
for parent, _ in nodes:
if "arguments" in parent:
arguments = parent["arguments"]
for argument in arguments:
variables.update(self.find_all_variables(argument))
return variables
def find_variables_in_if(self, node):
variables = []
def traverse(node):
if "condition" in node:
condition = node["condition"]
if (
"leftExpression" in condition
and condition["leftExpression"]["nodeType"] == "Identifier"
):
variables.append(condition["leftExpression"]["name"])
if (
"rightExpression" in condition
and condition["rightExpression"]["nodeType"] == "Identifier"
):
variables.append(condition["rightExpression"]["name"])
traverse(condition)
if "falseBody" in node and node["falseBody"]:
traverse(node["falseBody"])
if "trueBody" in node and node["trueBody"]:
if (
"nodeType" in node["trueBody"]
and node["trueBody"]["nodeType"] == "Block"
):
statements = node["trueBody"].get("statements", [])
for statement in statements:
traverse(statement)
else:
traverse(node["trueBody"])
if "body" in node and node["body"]:
if "nodeType" in node["body"] and node["body"]["nodeType"] == "Block":
statements = node["body"].get("statements", [])
for statement in statements:
traverse(statement)
else:
traverse(node["body"])
traverse(node)
return variables
def extract_address_variable(self, node):
if type(node) == int:
return set([])
transfer_vars = set([])
if (
node.get("nodeType", "") == "ExpressionStatement"
and node.get("expression", {}).get("nodeType") == "FunctionCall"
):
expression = node["expression"].get("expression", None)
if expression is not None:
if (
expression["nodeType"] == "MemberAccess"
and expression["memberName"] in TRANSFER_METHODS
):
print(expression)
address_variable = expression["expression"].get("name")
if address_variable:
transfer_vars.update(set([address_variable]))
for key, value in node.items():
if isinstance(value, dict):
transfer_vars.update(self.extract_address_variable(value))
elif isinstance(value, list):
for item in value:
transfer_vars.update(self.extract_address_variable(item))
return transfer_vars

@ -0,0 +1,72 @@
import pytest
from pathlib import Path
from mythril.mythril import MythrilDisassembler
from mythril.solidity.features import SolidityFeatureExtractor
from mythril.solidity.soliditycontract import SolidityContract, SolcAST
TEST_FILES = Path(__file__).parent / "testdata/input_contracts"
solc_binary = MythrilDisassembler._init_solc_binary("v0.5.0")
test_cases = [
("suicide.sol", 1, "kill", "contains_selfdestruct", True),
(
"SimpleModifier.sol",
1,
"withdrawfunds",
"all_require_vars",
set(["msg", "owner"]),
),
("SimpleModifier.sol", 1, "withdrawfunds", "transfer_vars", set(["owner"])),
("rubixi.sol", 18, "init", "contains_selfdestruct", False),
("rubixi.sol", 18, "collectAllFees", "has_owner_modifier", True),
("rubixi.sol", 18, "collectAllFees", "is_payable", False),
(
"rubixi.sol",
18,
"collectAllFees",
"all_require_vars",
set(["collectedFees", "creator"]),
),
(
"rubixi.sol",
18,
"collectPercentOfFees",
"all_require_vars",
set(["collectedFees", "_pcent", "creator"]),
),
(
"rubixi.sol",
18,
"changeMultiplier",
"all_require_vars",
set(["_mult", "creator"]),
),
("rubixi.sol", 18, "", "is_payable", True),
("rubixi.sol", 18, "collectAllFees", "transfer_vars", set(["creator"])),
("exceptions.sol", 8, "assert3", "contains_assert", True),
("exceptions.sol", 8, "requireisfine", "all_require_vars", set(["input"])),
("WalletLibrary.sol", 23, "execute", "has_owner_modifier", True),
("WalletLibrary.sol", 23, "initWallet", "has_owner_modifier", False),
("WalletLibrary.sol", 23, "initWallet", "all_require_vars", set(["m_numOwners"])),
("WalletLibrary.sol", 23, "confirm", "all_require_vars", set(["success"])),
("kcalls.sol", 3, "callSetN", "contains_call", True),
("kcalls.sol", 3, "delegatecallSetN", "contains_delegatecall", True),
("kcalls.sol", 3, "callcodeSetN", "contains_staticcall", True),
]
@pytest.mark.parametrize(
"file_name, num_funcs, func_name, field, expected_value", test_cases
)
def test_feature_selfdestruct(file_name, num_funcs, func_name, field, expected_value):
input_file = TEST_FILES / file_name
name = file_name.split(".")[0]
print(name, name.capitalize())
if name[0].islower():
name = name.capitalize()
contract = SolidityContract(str(input_file), name=name, solc_binary=solc_binary)
ms = contract.solc_json["sources"][str(input_file)]["ast"]
sfe = SolidityFeatureExtractor(ms)
fe = sfe.extract_features()
assert len(fe) == num_funcs
assert fe[func_name][field] == expected_value

@ -0,0 +1,19 @@
pragma solidity ^0.5.0;
contract SimpleModifier {
address payable public owner;
modifier onlyOwner() {
require(msg.sender == owner);
_;
}
function withdrawfunds() public onlyOwner {
owner.send(address(this).balance);
}
}

@ -0,0 +1,398 @@
//sol Wallet
// Multi-sig, daily-limited account proxy/wallet.
// @authors:
// Gav Wood <g@ethdev.com>
// inheritable "property" contract that enables methods to be protected by requiring the acquiescence of either a
// single, or, crucially, each of a number of, designated owners.
// usage:
// use modifiers onlyowner (just own owned) or onlymanyowners(hash), whereby the same hash must be provided by
// some number (specified in constructor) of the set of owners (specified in the constructor, modifiable) before the
// interior is executed.
pragma solidity 0.5.0;
contract WalletEvents {
// EVENTS
// this contract only has six types of events: it can accept a confirmation, in which case
// we record owner and operation (hash) alongside it.
event Confirmation(address owner, bytes32 operation);
event Revoke(address owner, bytes32 operation);
// some others are in the case of an owner changing.
event OwnerChanged(address oldOwner, address newOwner);
event OwnerAdded(address newOwner);
event OwnerRemoved(address oldOwner);
// the last one is emitted if the required signatures change
event RequirementChanged(uint newRequirement);
// Funds has arrived into the wallet (record how much).
event Deposit(address _from, uint value);
// Single transaction going out of the wallet (record who signed for it, how much, and to whom it's going).
event SingleTransact(address owner, uint value, address to, bytes data, address created);
// Multi-sig transaction going out of the wallet (record who signed for it last, the operation hash, how much, and to whom it's going).
event MultiTransact(address owner, bytes32 operation, uint value, address to, bytes data, address created);
// Confirmation still needed for a transaction.
event ConfirmationNeeded(bytes32 operation, address initiator, uint value, address to, bytes data);
}
contract WalletAbi {
// Revokes a prior confirmation of the given operation
function revoke(bytes32 _operation) external;
// Replaces an owner `_from` with another `_to`.
function changeOwner(address _from, address _to) external;
function addOwner(address _owner) external;
function removeOwner(address _owner) external;
function changeRequirement(uint _newRequired) external;
function isOwner(address _addr) public returns (bool);
function hasConfirmed(bytes32 _operation, address _owner) external returns (bool);
// (re)sets the daily limit. needs many of the owners to confirm. doesn't alter the amount already spent today.
function setDailyLimit(uint _newLimit) external;
function execute(address _to, uint _value, bytes calldata _data) external returns (bytes32 o_hash);
function confirm(bytes32 _h) public returns (bool o_success);
}
contract WalletLibrary is WalletEvents {
// TYPES
// struct for the status of a pending operation.
struct PendingState {
uint yetNeeded;
uint ownersDone;
uint index;
}
// Transaction structure to remember details of transaction lest it need be saved for a later call.
struct Transaction {
address to;
uint value;
bytes data;
}
// MODIFIERS
// simple single-sig function modifier.
modifier onlyowner {
if (isOwner(msg.sender))
_;
}
// multi-sig function modifier: the operation must have an intrinsic hash in order
// that later attempts can be realised as the same underlying operation and
// thus count as confirmations.
modifier onlymanyowners(bytes32 _operation) {
if (confirmAndCheck(_operation))
_;
}
// METHODS
// gets called when no other function matches
function() external payable {
// just being sent some cash?
if (msg.value > 0)
emit Deposit(msg.sender, msg.value);
}
// constructor is given number of sigs required to do protected "onlymanyowners" transactions
// as well as the selection of addresses capable of confirming them.
function initMultiowned(address[] memory _owners, uint _required) public only_uninitialized {
m_numOwners = _owners.length + 1;
m_owners[1] = uint(msg.sender);
m_ownerIndex[uint(msg.sender)] = 1;
for (uint i = 0; i < _owners.length; ++i)
{
m_owners[2 + i] = uint(_owners[i]);
m_ownerIndex[uint(_owners[i])] = 2 + i;
}
m_required = _required;
}
// Revokes a prior confirmation of the given operation
function revoke(bytes32 _operation) external {
uint ownerIndex = m_ownerIndex[uint(msg.sender)];
// make sure they're an owner
if (ownerIndex == 0) return;
uint ownerIndexBit = 2**ownerIndex;
PendingState memory pending = m_pending[_operation];
if (pending.ownersDone & ownerIndexBit > 0) {
pending.yetNeeded++;
pending.ownersDone -= ownerIndexBit;
emit Revoke(msg.sender, _operation);
}
}
// Replaces an owner `_from` with another `_to`.
function changeOwner(address _from, address _to) onlymanyowners(keccak256(msg.data)) external {
if (isOwner(_to)) return;
uint ownerIndex = m_ownerIndex[uint(_from)];
if (ownerIndex == 0) return;
clearPending();
m_owners[ownerIndex] = uint(_to);
m_ownerIndex[uint(_from)] = 0;
m_ownerIndex[uint(_to)] = ownerIndex;
emit OwnerChanged(_from, _to);
}
function addOwner(address _owner) onlymanyowners(keccak256(msg.data)) external {
if (isOwner(_owner)) return;
clearPending();
if (m_numOwners >= c_maxOwners)
reorganizeOwners();
if (m_numOwners >= c_maxOwners)
return;
m_numOwners++;
m_owners[m_numOwners] = uint(_owner);
m_ownerIndex[uint(_owner)] = m_numOwners;
emit OwnerAdded(_owner);
}
function removeOwner(address _owner) onlymanyowners(keccak256(msg.data)) external {
uint ownerIndex = m_ownerIndex[uint(_owner)];
if (ownerIndex == 0) return;
if (m_required > m_numOwners - 1) return;
m_owners[ownerIndex] = 0;
m_ownerIndex[uint(_owner)] = 0;
clearPending();
reorganizeOwners(); //make sure m_numOwner is equal to the number of owners and always points to the optimal free slot
emit OwnerRemoved(_owner);
}
function changeRequirement(uint _newRequired) onlymanyowners(keccak256(msg.data)) external {
if (_newRequired > m_numOwners) return;
m_required = _newRequired;
clearPending();
emit RequirementChanged(_newRequired);
}
// Gets an owner by 0-indexed position (using numOwners as the count)
function getOwner(uint ownerIndex) external view returns (address) {
return address(m_owners[ownerIndex + 1]);
}
function isOwner(address _addr) public view returns (bool) {
return m_ownerIndex[uint(_addr)] > 0;
}
function hasConfirmed(bytes32 _operation, address _owner) external view returns (bool) {
PendingState memory pending = m_pending[_operation];
uint ownerIndex = m_ownerIndex[uint(_owner)];
// make sure they're an owner
if (ownerIndex == 0) return false;
// determine the bit to set for this owner.
uint ownerIndexBit = 2**ownerIndex;
return !(pending.ownersDone & ownerIndexBit == 0);
}
// constructor - stores initial daily limit and records the present day's index.
function initDaylimit(uint _limit) public only_uninitialized {
m_dailyLimit = _limit;
m_lastDay = today();
}
// (re)sets the daily limit. needs many of the owners to confirm. doesn't alter the amount already spent today.
function setDailyLimit(uint _newLimit) onlymanyowners(keccak256(msg.data)) external {
m_dailyLimit = _newLimit;
}
// resets the amount already spent today. needs many of the owners to confirm.
function resetSpentToday() onlymanyowners(keccak256(msg.data)) external {
m_spentToday = 0;
}
// throw unless the contract is not yet initialized.
modifier only_uninitialized { require(m_numOwners == 0); _; }
// constructor - just pass on the owner array to the multiowned and
// the limit to daylimit
function initWallet(address[] memory _owners, uint _required, uint _daylimit) public only_uninitialized {
initDaylimit(_daylimit);
initMultiowned(_owners, _required);
}
// kills the contract sending everything to `_to`.
function kill(address payable _to) onlymanyowners(keccak256(msg.data)) external {
selfdestruct(_to);
}
// Outside-visible transact entry point. Executes transaction immediately if below daily spend limit.
// If not, goes into multisig process. We provide a hash on return to allow the sender to provide
// shortcuts for the other confirmations (allowing them to avoid replicating the _to, _value
// and _data arguments). They still get the option of using them if they want, anyways.
function execute(address _to, uint _value, bytes calldata _data) external onlyowner returns (bytes32 o_hash) {
// first, take the opportunity to check that we're under the daily limit.
if ((_data.length == 0 && underLimit(_value)) || m_required == 1) {
// yes - just execute the call.
address created;
if (_to == address(0)) {
created = create(_value, _data);
} else {
(bool success, bytes memory data) = _to.call.value(_value)(_data);
require(success);
}
emit SingleTransact(msg.sender, _value, _to, _data, created);
} else {
// determine our operation hash.
o_hash = keccak256(abi.encode(msg.data, block.number));
// store if it's new
if (m_txs[o_hash].to == address(0) && m_txs[o_hash].value == 0 && m_txs[o_hash].data.length == 0) {
m_txs[o_hash].to = _to;
m_txs[o_hash].value = _value;
m_txs[o_hash].data = _data;
}
if (!confirm(o_hash)) {
emit ConfirmationNeeded(o_hash, msg.sender, _value, _to, _data);
}
}
}
function create(uint _value, bytes memory _code) internal returns (address o_addr) {
uint256 o_size;
assembly {
o_addr := create(_value, add(_code, 0x20), mload(_code))
o_size := extcodesize(o_addr)
}
require(o_size != 0);
}
// confirm a transaction through just the hash. we use the previous transactions map, m_txs, in order
// to determine the body of the transaction from the hash provided.
function confirm(bytes32 _h) public onlymanyowners(_h) returns (bool o_success) {
if (m_txs[_h].to != address(0) || m_txs[_h].value != 0 || m_txs[_h].data.length != 0) {
address created;
if (m_txs[_h].to == address(0)) {
created = create(m_txs[_h].value, m_txs[_h].data);
} else {
(bool success, bytes memory data) = m_txs[_h].to.call.value(m_txs[_h].value)(m_txs[_h].data);
require(success);
}
emit MultiTransact(msg.sender, _h, m_txs[_h].value, m_txs[_h].to, m_txs[_h].data, created);
delete m_txs[_h];
return true;
}
}
// INTERNAL METHODS
function confirmAndCheck(bytes32 _operation) internal returns (bool) {
// determine what index the present sender is:
uint ownerIndex = m_ownerIndex[uint(msg.sender)];
// make sure they're an owner
if (ownerIndex == 0) return false;
PendingState memory pending = m_pending[_operation];
// if we're not yet working on this operation, switch over and reset the confirmation status.
if (pending.yetNeeded == 0) {
// reset count of confirmations needed.
pending.yetNeeded = m_required;
// reset which owners have confirmed (none) - set our bitmap to 0.
pending.ownersDone = 0;
pending.index = m_pendingIndex.length++;
m_pendingIndex[pending.index] = _operation;
}
// determine the bit to set for this owner.
uint ownerIndexBit = 2**ownerIndex;
// make sure we (the message sender) haven't confirmed this operation previously.
if (pending.ownersDone & ownerIndexBit == 0) {
emit Confirmation(msg.sender, _operation);
// ok - check if count is enough to go ahead.
if (pending.yetNeeded <= 1) {
// enough confirmations: reset and run interior.
delete m_pendingIndex[m_pending[_operation].index];
delete m_pending[_operation];
return true;
}
else
{
// not enough: record that this owner in particular confirmed.
pending.yetNeeded--;
pending.ownersDone |= ownerIndexBit;
}
}
}
function reorganizeOwners() private {
uint free = 1;
while (free < m_numOwners)
{
while (free < m_numOwners && m_owners[free] != 0) free++;
while (m_numOwners > 1 && m_owners[m_numOwners] == 0) m_numOwners--;
if (free < m_numOwners && m_owners[m_numOwners] != 0 && m_owners[free] == 0)
{
m_owners[free] = m_owners[m_numOwners];
m_ownerIndex[m_owners[free]] = free;
m_owners[m_numOwners] = 0;
}
}
}
// checks to see if there is at least `_value` left from the daily limit today. if there is, subtracts it and
// returns true. otherwise just returns false.
function underLimit(uint _value) internal onlyowner returns (bool) {
// reset the spend limit if we're on a different day to last time.
if (today() > m_lastDay) {
m_spentToday = 0;
m_lastDay = today();
}
// check to see if there's enough left - if so, subtract and return true.
// overflow protection // dailyLimit check
if (m_spentToday + _value >= m_spentToday && m_spentToday + _value <= m_dailyLimit) {
m_spentToday += _value;
return true;
}
return false;
}
// determines today's index.
function today() private view returns (uint) { return now / 1 days; }
function clearPending() internal {
uint length = m_pendingIndex.length;
for (uint i = 0; i < length; ++i) {
delete m_txs[m_pendingIndex[i]];
if (m_pendingIndex[i] != 0)
delete m_pending[m_pendingIndex[i]];
}
delete m_pendingIndex;
}
// FIELDS
address _walletLibrary = 0xCAfEcAfeCAfECaFeCaFecaFecaFECafECafeCaFe;
// the number of owners that must confirm the same operation before it is run.
uint public m_required;
// pointer used to find a free slot in m_owners
uint public m_numOwners;
uint public m_dailyLimit;
uint public m_spentToday;
uint public m_lastDay;
// list of owners
uint[256] m_owners;
uint c_maxOwners = 250;
// index on the list of owners to allow reverse lookup
mapping(uint => uint) m_ownerIndex;
// the ongoing operations.
mapping(bytes32 => PendingState) m_pending;
bytes32[] m_pendingIndex;
// pending transactions we have at present.
mapping (bytes32 => Transaction) m_txs;
}

@ -16,7 +16,9 @@ contract Exceptions {
}
function assert3(uint256 input) public pure {
assert(input != 23);
if (input > 10) {
assert(input != 23);
}
}
function requireisfine(uint256 input) public pure {

@ -1,7 +1,7 @@
pragma solidity ^0.5.0;
contract D {
contract Kcalls {
uint public n;
address public sender;
Loading…
Cancel
Save