Better tracing

contractstorage.py

@@ -1,8 +1,10 @@
 from rpc.client import EthJsonRpc
 from ethcontract import ETHContract
+from ethereum import utils
 from tinydb import TinyDB, Query
 import codecs
 import hashlib
+import re
 
 
 class ContractStorage:
@@ -78,13 +80,23 @@ class ContractStorage:
 
         all_contracts = self.db.all()
 
+        matches = re.findall(r'func\[([a-zA-Z0-9\s,()]+)\]', expression)
+
+        for m in matches:
+            # Pre-calculate function signature hashes
+
+            sign_hash = utils.sha3(m)[:4].hex()
+
+            print(sign_hash)
+
+            expression = expression.replace(m, sign_hash)
+
         for c in all_contracts:
 
             for instance in c['instances']:
 
-                if ('balance' in instance):
-                    contract = ETHContract(c['code'], instance['balance'])
+                contract = ETHContract(c['code'], instance['balance'])
 
-                    if (contract.matches_expression(expression)):
-                        print("Found contract:" + instance['address'])
+                if (contract.matches_expression(expression)):
+                    print("Found contract:" + instance['address'])
 

ethcontract.py

@@ -2,14 +2,15 @@ from ether import asm, util
 import re
 
 
-
 class ETHContract:
 
     def __init__(self, code = "", balance = 0):
 
-        self.easm_code = asm.disassembly_to_easm(asm.disassemble(util.safe_decode(code)))
+        self.disassembly = asm.disassemble(util.safe_decode(code))
+        self.easm_code = asm.disassembly_to_easm(self.disassembly)
         self.balance = balance
 
+
     def matches_expression(self, expression):
 
         str_eval = ""
@@ -22,7 +23,7 @@ class ETHContract:
                 str_eval += token
                 continue
 
-            m = re.match(r'^code\(([a-zA-Z0-9\s,]+)\)$', token)
+            m = re.match(r'^code\[([a-zA-Z0-9\s,]+)\]$', token)
 
             if (m):
                 code = m.group(1).replace(",", "\\n")
@@ -30,9 +31,17 @@ class ETHContract:
                 continue
 
             m = re.match(r'^balance\s*[=><]+\s*\d+$', token)
+
             if (m):
                 str_eval += "self." + m.group(0)
                 continue
 
+            m = re.match(r'^func\[([a-zA-Z0-9\s,()]+)\]$', token)
+
+            if (m):
+                str_eval += "\"" + m.group(1) + "\" in self.easm_code"               
+
+                continue
+
         return eval(str_eval)
 

ether/asm.py

@@ -83,23 +83,6 @@ def find_opcode_sequence(pattern, disassembly):
     return match_indexes
 
 
-def resolve_functions(disassembly):
-
-    function_stubs = find_opcode_sequence(['PUSH4', 'EQ', 'PUSH2', 'JUMPI'], disassembly)
-
-    functions = []
-
-    for index in function_stubs:
-        func = {}
-
-        func['hash'] = disassembly[index]['argument']
-        func['address'] = disassembly[index + 2]['argument']
-
-        functions.append(func)
-
-    return functions
-
-
 def disassemble(bytecode):
 
     disassembly = []

ether/evm.py

@@ -4,9 +4,10 @@ from ethereum.slogging import get_logger
 from logging import StreamHandler
 import sys
 import codecs
+from .util import safe_decode
 
 
-def trace(code):
+def trace(code, address = "", calldata = ""):
 
 	logHandlers = ['eth.vm.op', 'eth.vm.op.stack', 'eth.vm.op.memory', 'eth.vm.op.storage']
 
@@ -17,12 +18,14 @@ def trace(code):
 		log_vm_op.setLevel("TRACE")
 		log_vm_op.addHandler(streamHandler)
 
-	addr = codecs.decode('0123456789ABCDEF0123456789ABCDEF01234567', 'hex_codec')
+	addr_from = codecs.decode('0123456789ABCDEF0123456789ABCDEF01234567', 'hex_codec')
+	addr_to = safe_decode(address)
+	data = safe_decode(calldata)
 
 	state = State()
 
-	ext = messages.VMExt(state, transactions.Transaction(0, 0, 21000, addr, 0, addr))
+	ext = messages.VMExt(state, transactions.Transaction(0, 0, 21000, addr_from, 0, addr_to))
 
-	msg = vm.Message(addr, addr)
+	message = vm.Message(addr_from, addr_to, 0, 21000, data, code_address=addr_to)
 
-	vm.vm_execute(ext, msg, code)
+	res, gas, dat = vm.vm_execute(ext, message, code)

mythril

@@ -1,5 +1,5 @@
 #!/usr/bin/env python
-"""mythril.py: Smart contract bug hunting on the Ethereum blockchain
+"""mythril.py: Bug hunting on the Ethereum blockchain
 
    http://www.github.com/b-mueller/mythril
 """
@@ -7,8 +7,8 @@
 from ether import asm,evm,util
 from contractstorage import ContractStorage
 import sys
-import codecs
 import argparse
+from rpc.client import EthJsonRpc
 
 
 def exitWithError(message):
@@ -16,15 +16,16 @@ def exitWithError(message):
     sys.exit()
 
 
-parser = argparse.ArgumentParser(description='Smart contract bug hunting on the Ethereum blockchain')
+parser = argparse.ArgumentParser(description='Bug hunting on the Ethereum blockchain')
 
-parser.add_argument('-d', '--disassemble',  action='store_true', help='disassemble, use with -c or --id')
-parser.add_argument('-t', '--trace',  action='store_true', help='trace bytecode provided via the -c argument')
+parser.add_argument('-d', '--disassemble',  action='store_true', help='disassemble, use with -c or -a')
+parser.add_argument('-t', '--trace', action='store_true', help='trace, use with -c or -a and --data (optional)')
 parser.add_argument('-c', '--code', help='hex-encoded bytecode string ("6060604052...")', metavar='BYTECODE')
+parser.add_argument('-a', '--address', default='0x0123456789ABCDEF0123456789ABCDEF01234567', help='contract address')
 parser.add_argument('-o', '--outfile')
-parser.add_argument('--id', help='contract id')
+parser.add_argument('--data', help='message call input data for tracing')
+parser.add_argument('--search', help='search the contract database')
 parser.add_argument('--init-db', action='store_true', help='Initialize the contract database')
-parser.add_argument('--search', help='search contract database')
 parser.add_argument('--rpchost', default='127.0.0.1', help='RPC host')
 parser.add_argument('--rpcport', type=int, default=8545, help='RPC port')
 
@@ -38,10 +39,14 @@ if (args.disassemble):
 
     if (args.code):
         encoded_bytecode = args.code
-    elif (args.id):
+    elif (args.address):
 
         try:
-            encoded_bytecode = storage.get_contract_code_by_address(args.id)
+            # encoded_bytecode = storage.get_contract_code_by_address(args.id)
+
+            eth = EthJsonRpc(args.rpchost, args.rpcport)
+
+            encoded_bytecode = eth.eth_getCode(args.address)
 
         except Exception as e:
             exitWithError("Exception loading bytecode via RPC: " + str(e))
@@ -68,14 +73,24 @@ if (args.disassemble):
 
 elif (args.trace):
 
-    if args.code:     
+    if (args.code):
+        encoded_bytecode = args.code
+
+    elif (args.address):
+
+        eth = EthJsonRpc(args.rpchost, args.rpcport)
+
+        encoded_bytecode = eth.eth_getCode(args.address)
+
+    else:
+        exitWithError("Disassembler: Provide the input bytecode via -c BYTECODE or --id ID")
 
-        bytecode = util.safe_decode(args.code)
+    if (args.data):
+        evm.trace(util.safe_decode(encoded_bytecode), args.address, args.data)
 
     else:
-        exitWithError("Trace: Provide the input bytecode using -c BYTECODE or -f INPUT_FILE")
+        evm.trace(util.safe_decode(encoded_bytecode), args.address)
 
-    evm.trace(bytecode)
 
 elif (args.search):
 

utils/download_sigs.py

@@ -0,0 +1,29 @@
+import requests
+from requests.adapters import HTTPAdapter
+
+JSON_MEDIA_TYPE = 'application/json'
+MAX_RETRIES = 3
+
+
+
+
+session = requests.Session()
+session.mount("www.4byte.directory", HTTPAdapter(max_retries=MAX_RETRIES))
+
+result_list = []
+headers = {'Content-Type': JSON_MEDIA_TYPE}
+
+for page in range(1,60):
+    print("retrieving page " + str(page))
+
+    url = 'https://www.4byte.directory/api/v1/signatures/?format=json&page=' + str(page)
+
+    try:
+        r = session.get(url, headers=headers)
+        response = r.json()
+    except ValueError:
+        raise BadJsonError(r.text)
+
+    result_list += response['results']
+
+print(result_list)