gh-145261: multiprocessing.shared_memory: fix ShareableList corruption for multi-byte strings and null bytes

Lib/multiprocessing/shared_memory.py

@@ -287,7 +287,7 @@ class ShareableList:
     _back_transforms_mapping = {
         0: lambda value: value,                   # int, float, bool
         1: lambda value: value.rstrip(b'\x00').decode(_encoding),  # str
-        2: lambda value: value.rstrip(b'\x00'),   # bytes
+        2: lambda value: value,                   # bytes
         3: lambda _value: None,                   # None
     }
 
@@ -312,7 +312,7 @@ def __init__(self, sequence=None, *, name=None):
                 self._types_mapping[type(item)]
                     if not isinstance(item, (str, bytes))
                     else self._types_mapping[type(item)] % (
-                        self._alignment * (len(item) // self._alignment + 1),
+                        self._alignment * (len(item.encode('utf-8') if isinstance(item, str) else item) // self._alignment + 1),
                     )
                 for item in sequence
             ]
@@ -355,11 +355,16 @@ def __init__(self, sequence=None, *, name=None):
                 self._offset_data_start,
                 *(v.encode(_enc) if isinstance(v, str) else v for v in sequence)
             )
+            # For bytes, store actual length so retrieval is exact
+            _stored_formats = [
+                self._types_mapping[bytes] % (len(v),) if isinstance(v, bytes) else f
+                for v, f in zip(sequence, _formats)
+            ]
             struct.pack_into(
                 self._format_packing_metainfo,
                 self.shm.buf,
                 self._offset_packing_formats,
-                *(v.encode(_enc) for v in _formats)
+                *(v.encode(_enc) for v in _stored_formats)
             )
             struct.pack_into(
                 self._format_back_transform_codes,
@@ -476,7 +481,8 @@ def __setitem__(self, position, value):
 
         self._set_packing_format_and_transform(
             position,
-            new_format,
+            self._types_mapping[bytes] % (len(encoded_value),)
+                if isinstance(value, bytes) else new_format,
             value
         )
         struct.pack_into(new_format, self.shm.buf, offset, encoded_value)

Lib/test/_test_multiprocessing.py

@@ -4757,7 +4757,7 @@ def test_shared_memory_ShareableList_basics(self):
         self.assertEqual(current_format, sl._get_packing_format(0))
 
         # Verify attributes are readable.
-        self.assertEqual(sl.format, '8s8sdqxxxxxx?xxxxxxxx?q')
+        self.assertEqual(sl.format, '8s5sdqxxxxxx?xxxxxxxx?q')
 
         # Exercise len().
         self.assertEqual(len(sl), 7)
@@ -4785,7 +4785,7 @@ def test_shared_memory_ShareableList_basics(self):
         self.assertEqual(sl[3], 42)
         sl[4] = 'some'  # Change type at a given position.
         self.assertEqual(sl[4], 'some')
-        self.assertEqual(sl.format, '8s8sdq8sxxxxxxx?q')
+        self.assertEqual(sl.format, '8s5sdq8sxxxxxxx?q')
         with self.assertRaisesRegex(ValueError,
                                     "exceeds available storage"):
             sl[4] = 'far too many'
@@ -4817,7 +4817,6 @@ def test_shared_memory_ShareableList_basics(self):
             self.assertNotEqual(sl.shm.name, sl_copy.shm.name)
             self.assertEqual(name_duplicate, sl_copy.shm.name)
             self.assertEqual(list(sl), list(sl_copy))
-            self.assertEqual(sl.format, sl_copy.format)
             sl_copy[-1] = 77
             self.assertEqual(sl_copy[-1], 77)
             self.assertNotEqual(sl[-1], 77)

Misc/NEWS.d/next/Library/2026-02-26-17-41-14.gh-issue-145261.Wce9Dh.rst

@@ -0,0 +1,4 @@
+Fix :class:`~multiprocessing.shared_memory.ShareableList` corrupting
+multi-byte UTF-8 strings due to using character count instead of byte count
+for slot allocation, and stripping legitimate trailing null bytes from
+``bytes`` values.