gh-112625: Protect bytearray from being freed by misbehaving iterator inside bytearra...

diff --git a/Lib/test/test_builtin.py b/Lib/test/test_builtin.py
index b7966f8f03875b301a2e9ad3c3d724b72bc7d23a..535856adaea4d353cebe45758ecea67fb064b3d4 100644 (file)
--- a/Lib/test/test_builtin.py
+++ b/Lib/test/test_builtin.py
@@ -2039,6 +2039,23 @@ class BuiltinTest(unittest.TestCase):
         bad_iter = map(int, "X")
         self.assertRaises(ValueError, array.extend, bad_iter)
 
+    def test_bytearray_join_with_misbehaving_iterator(self):
+        # Issue #112625
+        array = bytearray(b',')
+        def iterator():
+            array.clear()
+            yield b'A'
+            yield b'B'
+        self.assertRaises(BufferError, array.join, iterator())
+
+    def test_bytearray_join_with_custom_iterator(self):
+        # Issue #112625
+        array = bytearray(b',')
+        def iterator():
+            yield b'A'
+            yield b'B'
+        self.assertEqual(bytearray(b'A,B'), array.join(iterator()))
+
     def test_construct_singletons(self):
         for const in None, Ellipsis, NotImplemented:
             tp = type(const)

diff --git a/Misc/NEWS.d/next/Core and Builtins/2023-12-03-19-34-51.gh-issue-112625.QWTlwS.rst b/Misc/NEWS.d/next/Core and Builtins/2023-12-03-19-34-51.gh-issue-112625.QWTlwS.rst
new file mode 100644 (file)
index 0000000..4970e10
--- /dev/null
+++ b/Misc/NEWS.d/next/Core and Builtins/2023-12-03-19-34-51.gh-issue-112625.QWTlwS.rst
@@ -0,0 +1 @@
+Fixes a bug where a bytearray object could be cleared while iterating over an argument in the ``bytearray.join()`` method that could result in reading memory after it was freed.

diff --git a/Objects/bytearrayobject.c b/Objects/bytearrayobject.c
index 67073190cc889d1fd29c25e9f8f240a99e8d339f..659de7d3dd5a9942ccc1ff8698ca59794e2686f9 100644 (file)
--- a/Objects/bytearrayobject.c
+++ b/Objects/bytearrayobject.c
@@ -2007,7 +2007,10 @@ static PyObject *
 bytearray_join(PyByteArrayObject *self, PyObject *iterable_of_bytes)
 /*[clinic end generated code: output=a8516370bf68ae08 input=aba6b1f9b30fcb8e]*/
 {
-    return stringlib_bytes_join((PyObject*)self, iterable_of_bytes);
+    self->ob_exports++; // this protects `self` from being cleared/resized if `iterable_of_bytes` is a custom iterator
+    PyObject* ret = stringlib_bytes_join((PyObject*)self, iterable_of_bytes);
+    self->ob_exports--; // unexport `self`
+    return ret;
 }
 
 /*[clinic input]