GH-95245: Move weakreflist into the pre-header. (GH-95996)

Author: markshannon

Include/internal/pycore_object.h

@@ -279,7 +279,7 @@ static inline size_t
 _PyType_PreHeaderSize(PyTypeObject *tp)
 {
     return _PyType_IS_GC(tp) * sizeof(PyGC_Head) +
-        _PyType_HasFeature(tp, Py_TPFLAGS_MANAGED_DICT) * 2 * sizeof(PyObject *);
+        _PyType_HasFeature(tp, Py_TPFLAGS_PREHEADER) * 2 * sizeof(PyObject *);
 }
 
 void _PyObject_GC_Link(PyObject *op);
@@ -296,7 +296,7 @@ extern int _Py_CheckSlotResult(
 
 // Test if a type supports weak references
 static inline int _PyType_SUPPORTS_WEAKREFS(PyTypeObject *type) {
-    return (type->tp_weaklistoffset > 0);
+    return (type->tp_weaklistoffset != 0);
 }
 
 extern PyObject* _PyType_AllocNoTrack(PyTypeObject *type, Py_ssize_t nitems);
@@ -346,7 +346,7 @@ _PyDictOrValues_SetValues(PyDictOrValues *ptr, PyDictValues *values)
     ptr->values = ((char *)values) - 1;
 }
 
-#define MANAGED_DICT_OFFSET (((int)sizeof(PyObject *))*-3)
+#define MANAGED_WEAKREF_OFFSET (((Py_ssize_t)sizeof(PyObject *))*-4)
 
 extern PyObject ** _PyObject_ComputedDictPointer(PyObject *);
 extern void _PyObject_FreeInstanceAttributes(PyObject *obj);

Include/object.h

@@ -360,12 +360,18 @@ given type object has a specified feature.
 /* Track types initialized using _PyStaticType_InitBuiltin(). */
 #define _Py_TPFLAGS_STATIC_BUILTIN (1 << 1)
 
+/* Placement of weakref pointers are managed by the VM, not by the type.
+ * The VM will automatically set tp_weaklistoffset.
+ */
+#define Py_TPFLAGS_MANAGED_WEAKREF (1 << 3)
+
 /* Placement of dict (and values) pointers are managed by the VM, not by the type.
- * The VM will automatically set tp_dictoffset. Should not be used for variable sized
- * classes, such as classes that extend tuple.
+ * The VM will automatically set tp_dictoffset.
  */
 #define Py_TPFLAGS_MANAGED_DICT (1 << 4)
 
+#define Py_TPFLAGS_PREHEADER (Py_TPFLAGS_MANAGED_WEAKREF | Py_TPFLAGS_MANAGED_DICT)
+
 /* Set if instances of the type object are treated as sequences for pattern matching */
 #define Py_TPFLAGS_SEQUENCE (1 << 5)
 /* Set if instances of the type object are treated as mappings for pattern matching */

Lib/test/test_capi.py

@@ -579,6 +579,37 @@ def test_heaptype_with_weakref(self):
         self.assertEqual(ref(), inst)
         self.assertEqual(inst.weakreflist, ref)
 
+    def test_heaptype_with_managed_weakref(self):
+        inst = _testcapi.HeapCTypeWithManagedWeakref()
+        ref = weakref.ref(inst)
+        self.assertEqual(ref(), inst)
+
+    def test_sublclassing_managed_weakref(self):
+
+        class C(_testcapi.HeapCTypeWithManagedWeakref):
+            pass
+
+        inst = C()
+        ref = weakref.ref(inst)
+        self.assertEqual(ref(), inst)
+
+    def test_sublclassing_managed_both(self):
+
+        class C1(_testcapi.HeapCTypeWithManagedWeakref, _testcapi.HeapCTypeWithManagedDict):
+            pass
+
+        class C2(_testcapi.HeapCTypeWithManagedDict, _testcapi.HeapCTypeWithManagedWeakref):
+            pass
+
+        for cls in (C1, C2):
+            inst = cls()
+            ref = weakref.ref(inst)
+            self.assertEqual(ref(), inst)
+            inst.spam = inst
+            del inst
+            ref = weakref.ref(cls())
+            self.assertIs(ref(), None)
+
     def test_heaptype_with_buffer(self):
         inst = _testcapi.HeapCTypeWithBuffer()
         b = bytes(inst)

Misc/NEWS.d/next/Core and Builtins/2022-08-15-12-41-14.gh-issue-95245.N4gOUV.rst

@@ -0,0 +1,3 @@
+Reduces the size of a "simple" Python object from 8 to 6 words by moving the
+weakreflist pointer into the pre-header directly before the object's
+dict/values pointer.

Modules/_testcapi/heaptype.c

@@ -801,6 +801,32 @@ static PyType_Spec  HeapCTypeWithManagedDict_spec = {
     HeapCTypeWithManagedDict_slots
 };
 
+static void
+heapctypewithmanagedweakref_dealloc(PyObject* self)
+{
+
+    PyTypeObject *tp = Py_TYPE(self);
+    PyObject_ClearWeakRefs(self);
+    PyObject_GC_UnTrack(self);
+    PyObject_GC_Del(self);
+    Py_DECREF(tp);
+}
+
+static PyType_Slot HeapCTypeWithManagedWeakref_slots[] = {
+    {Py_tp_traverse, heapgcctype_traverse},
+    {Py_tp_getset, heapctypewithdict_getsetlist},
+    {Py_tp_dealloc, heapctypewithmanagedweakref_dealloc},
+    {0, 0},
+};
+
+static PyType_Spec  HeapCTypeWithManagedWeakref_spec = {
+    "_testcapi.HeapCTypeWithManagedWeakref",
+    sizeof(PyObject),
+    0,
+    Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE | Py_TPFLAGS_HAVE_GC | Py_TPFLAGS_MANAGED_WEAKREF,
+    HeapCTypeWithManagedWeakref_slots
+};
+
 static struct PyMemberDef heapctypewithnegativedict_members[] = {
     {"dictobj", T_OBJECT, offsetof(HeapCTypeWithDictObject, dict)},
     {"__dictoffset__", T_PYSSIZET, -(Py_ssize_t)sizeof(void*), READONLY},
@@ -1009,6 +1035,12 @@ _PyTestCapi_Init_Heaptype(PyObject *m) {
     }
     PyModule_AddObject(m, "HeapCTypeWithManagedDict", HeapCTypeWithManagedDict);
 
+    PyObject *HeapCTypeWithManagedWeakref = PyType_FromSpec(&HeapCTypeWithManagedWeakref_spec);
+    if (HeapCTypeWithManagedWeakref == NULL) {
+        return -1;
+    }
+    PyModule_AddObject(m, "HeapCTypeWithManagedWeakref", HeapCTypeWithManagedWeakref);
+
     PyObject *HeapCTypeWithWeakref = PyType_FromSpec(&HeapCTypeWithWeakref_spec);
     if (HeapCTypeWithWeakref == NULL) {
         return -1;

Objects/typeobject.c

@@ -2497,10 +2497,11 @@ subtype_getweakref(PyObject *obj, void *context)
         return NULL;
     }
     _PyObject_ASSERT((PyObject *)type,
-                     type->tp_weaklistoffset > 0);
+                     type->tp_weaklistoffset > 0 ||
+                     type->tp_weaklistoffset == MANAGED_WEAKREF_OFFSET);
     _PyObject_ASSERT((PyObject *)type,
-                     ((type->tp_weaklistoffset + sizeof(PyObject *))
-                      <= (size_t)(type->tp_basicsize)));
+                     ((type->tp_weaklistoffset + (Py_ssize_t)sizeof(PyObject *))
+                      <= type->tp_basicsize));
     weaklistptr = (PyObject **)((char *)obj + type->tp_weaklistoffset);
     if (*weaklistptr == NULL)
         result = Py_None;
@@ -3093,9 +3094,9 @@ type_new_descriptors(const type_new_ctx *ctx, PyTypeObject *type)
     }
 
     if (ctx->add_weak) {
-        assert(!ctx->base->tp_itemsize);
-        type->tp_weaklistoffset = slotoffset;
-        slotoffset += sizeof(PyObject *);
+        assert((type->tp_flags & Py_TPFLAGS_MANAGED_WEAKREF) == 0);
+        type->tp_flags |= Py_TPFLAGS_MANAGED_WEAKREF;
+        type->tp_weaklistoffset = MANAGED_WEAKREF_OFFSET;
     }
     if (ctx->add_dict) {
         assert((type->tp_flags & Py_TPFLAGS_MANAGED_DICT) == 0);
@@ -5116,9 +5117,9 @@ compatible_for_assignment(PyTypeObject* oldto, PyTypeObject* newto, const char*
          !same_slots_added(newbase, oldbase))) {
         goto differs;
     }
-    /* The above does not check for managed __dicts__ */
-    if ((oldto->tp_flags & Py_TPFLAGS_MANAGED_DICT) ==
-        ((newto->tp_flags & Py_TPFLAGS_MANAGED_DICT)))
+    /* The above does not check for the preheader */
+    if ((oldto->tp_flags & Py_TPFLAGS_PREHEADER) ==
+        ((newto->tp_flags & Py_TPFLAGS_PREHEADER)))
     {
         return 1;
     }
@@ -5217,7 +5218,7 @@ object_set_class(PyObject *self, PyObject *value, void *closure)
     if (compatible_for_assignment(oldto, newto, "__class__")) {
         /* Changing the class will change the implicit dict keys,
          * so we must materialize the dictionary first. */
-        assert((oldto->tp_flags & Py_TPFLAGS_MANAGED_DICT) == (newto->tp_flags & Py_TPFLAGS_MANAGED_DICT));
+        assert((oldto->tp_flags & Py_TPFLAGS_PREHEADER) == (newto->tp_flags & Py_TPFLAGS_PREHEADER));
         _PyObject_GetDictPtr(self);
         if (oldto->tp_flags & Py_TPFLAGS_MANAGED_DICT &&
             _PyDictOrValues_IsValues(*_PyObject_DictOrValuesPointer(self)))
@@ -5360,7 +5361,7 @@ object_getstate_default(PyObject *obj, int required)
         {
             basicsize += sizeof(PyObject *);
         }
-        if (Py_TYPE(obj)->tp_weaklistoffset) {
+        if (Py_TYPE(obj)->tp_weaklistoffset > 0) {
             basicsize += sizeof(PyObject *);
         }
         if (slotnames != Py_None) {
@@ -6150,7 +6151,7 @@ inherit_special(PyTypeObject *type, PyTypeObject *base)
         if (type->tp_clear == NULL)
             type->tp_clear = base->tp_clear;
     }
-    type->tp_flags |= (base->tp_flags & Py_TPFLAGS_MANAGED_DICT);
+    type->tp_flags |= (base->tp_flags & Py_TPFLAGS_PREHEADER);
 
     if (type->tp_basicsize == 0)
         type->tp_basicsize = base->tp_basicsize;
@@ -6571,7 +6572,7 @@ type_ready_fill_dict(PyTypeObject *type)
 }
 
 static int
-type_ready_dict_offset(PyTypeObject *type)
+type_ready_preheader(PyTypeObject *type)
 {
     if (type->tp_flags & Py_TPFLAGS_MANAGED_DICT) {
         if (type->tp_dictoffset > 0 || type->tp_dictoffset < -1) {
@@ -6583,6 +6584,18 @@ type_ready_dict_offset(PyTypeObject *type)
         }
         type->tp_dictoffset = -1;
     }
+    if (type->tp_flags & Py_TPFLAGS_MANAGED_WEAKREF) {
+        if (type->tp_weaklistoffset != 0 &&
+            type->tp_weaklistoffset != MANAGED_WEAKREF_OFFSET)
+        {
+            PyErr_Format(PyExc_TypeError,
+                        "type %s has the Py_TPFLAGS_MANAGED_WEAKREF flag "
+                        "but tp_weaklistoffset is set",
+                        type->tp_name);
+            return -1;
+        }
+        type->tp_weaklistoffset = MANAGED_WEAKREF_OFFSET;
+    }
     return 0;
 }
 
@@ -6802,7 +6815,7 @@ type_ready_post_checks(PyTypeObject *type)
             return -1;
         }
     }
-    else if (type->tp_dictoffset < sizeof(PyObject)) {
+    else if (type->tp_dictoffset < (Py_ssize_t)sizeof(PyObject)) {
         if (type->tp_dictoffset + type->tp_basicsize <= 0) {
             PyErr_Format(PyExc_SystemError,
                          "type %s has a tp_dictoffset that is too small");
@@ -6847,7 +6860,7 @@ type_ready(PyTypeObject *type)
     if (type_ready_inherit(type) < 0) {
         return -1;
     }
-    if (type_ready_dict_offset(type) < 0) {
+    if (type_ready_preheader(type) < 0) {
         return -1;
     }
     if (type_ready_set_hash(type) < 0) {