Issue #5308: Raise ValueError when marshalling too large object (a sequence

with size >= 2**31), instead of producing illegal marshal data.

Lib/test/test_marshal.py

@@ -269,6 +269,53 @@ class BugsTestCase(unittest.TestCase):
         invalid_string = 'l\x02\x00\x00\x00\x00\x00\x00\x00'
         self.assertRaises(ValueError, marshal.loads, invalid_string)
 
+LARGE_SIZE = 2**31
+character_size = 4 if sys.maxunicode > 0xFFFF else 2
+pointer_size = 8 if sys.maxsize > 0xFFFFFFFF else 4
+
+@unittest.skipIf(LARGE_SIZE > sys.maxsize, "test cannot run on 32-bit systems")
+class LargeValuesTestCase(unittest.TestCase):
+    def check_unmarshallable(self, data):
+        f = open(test_support.TESTFN, 'wb')
+        self.addCleanup(test_support.unlink, test_support.TESTFN)
+        with f:
+            self.assertRaises(ValueError, marshal.dump, data, f)
+
+    @test_support.precisionbigmemtest(size=LARGE_SIZE, memuse=1, dry_run=False)
+    def test_bytes(self, size):
+        self.check_unmarshallable(b'x' * size)
+
+    @test_support.precisionbigmemtest(size=LARGE_SIZE,
+            memuse=character_size, dry_run=False)
+    def test_str(self, size):
+        self.check_unmarshallable('x' * size)
+
+    @test_support.precisionbigmemtest(size=LARGE_SIZE,
+            memuse=pointer_size, dry_run=False)
+    def test_tuple(self, size):
+        self.check_unmarshallable((None,) * size)
+
+    @test_support.precisionbigmemtest(size=LARGE_SIZE,
+            memuse=pointer_size, dry_run=False)
+    def test_list(self, size):
+        self.check_unmarshallable([None] * size)
+
+    @test_support.precisionbigmemtest(size=LARGE_SIZE,
+            memuse=pointer_size*12 + sys.getsizeof(LARGE_SIZE-1),
+            dry_run=False)
+    def test_set(self, size):
+        self.check_unmarshallable(set(range(size)))
+
+    @test_support.precisionbigmemtest(size=LARGE_SIZE,
+            memuse=pointer_size*12 + sys.getsizeof(LARGE_SIZE-1),
+            dry_run=False)
+    def test_frozenset(self, size):
+        self.check_unmarshallable(frozenset(range(size)))
+
+    @test_support.precisionbigmemtest(size=LARGE_SIZE, memuse=1, dry_run=False)
+    def test_bytearray(self, size):
+        self.check_unmarshallable(bytearray(size))
+
 
 def test_main():
     test_support.run_unittest(IntTestCase,
@@ -277,7 +324,9 @@ def test_main():
                               CodeTestCase,
                               ContainerTestCase,
                               ExceptionTestCase,
-                              BugsTestCase)
+                              BugsTestCase,
+                              LargeValuesTestCase,
+                             )
 
 if __name__ == "__main__":
     test_main()

Lib/test/test_support.py

@@ -1062,7 +1062,7 @@ def bigmemtest(minsize, memuse, overhead=5*_1M):
         return wrapper
     return decorator
 
-def precisionbigmemtest(size, memuse, overhead=5*_1M):
+def precisionbigmemtest(size, memuse, overhead=5*_1M, dry_run=True):
     def decorator(f):
         def wrapper(self):
             if not real_max_memuse:
@@ -1070,11 +1070,12 @@ def precisionbigmemtest(size, memuse, overhead=5*_1M):
             else:
                 maxsize = size
 
-                if real_max_memuse and real_max_memuse < maxsize * memuse:
-                    if verbose:
-                        sys.stderr.write("Skipping %s because of memory "
-                                         "constraint\n" % (f.__name__,))
-                    return
+            if ((real_max_memuse or not dry_run)
+                and real_max_memuse < maxsize * memuse):
+                if verbose:
+                    sys.stderr.write("Skipping %s because of memory "
+                                     "constraint\n" % (f.__name__,))
+                return
 
             return f(self, maxsize)
         wrapper.size = size

Misc/NEWS

@@ -9,6 +9,9 @@ What's New in Python 2.7.4
 Core and Builtins
 -----------------
 
+- Issue #5308: Raise ValueError when marshalling too large object (a sequence
+  with size >= 2**31), instead of producing illegal marshal data.
+
 - Issue #17043: The unicode-internal decoder no longer read past the end of
   input buffer.