diff options
author | Dino Viehland <dinoviehland@meta.com> | 2024-04-22 05:57:05 (GMT) |
---|---|---|
committer | GitHub <noreply@github.com> | 2024-04-22 05:57:05 (GMT) |
commit | 8b541c017ea92040add608b3e0ef8dc85e9e6060 (patch) | |
tree | 880ff7c97041f374d21197b1b486686f2d64c998 | |
parent | 1446024124fb98c3051199760380685f8a2fd127 (diff) | |
download | cpython-8b541c017ea92040add608b3e0ef8dc85e9e6060.zip cpython-8b541c017ea92040add608b3e0ef8dc85e9e6060.tar.gz cpython-8b541c017ea92040add608b3e0ef8dc85e9e6060.tar.bz2 |
gh-112075: Make instance attributes stored in inline "dict" thread safe (#114742)
Make instance attributes stored in inline "dict" thread safe on free-threaded builds
-rw-r--r-- | Include/cpython/object.h | 1 | ||||
-rw-r--r-- | Include/internal/pycore_dict.h | 19 | ||||
-rw-r--r-- | Include/internal/pycore_object.h | 16 | ||||
-rw-r--r-- | Include/internal/pycore_pyatomic_ft_wrappers.h | 14 | ||||
-rw-r--r-- | Lib/test/test_class.py | 9 | ||||
-rw-r--r-- | Objects/dictobject.c | 385 | ||||
-rw-r--r-- | Objects/object.c | 45 | ||||
-rw-r--r-- | Objects/typeobject.c | 30 | ||||
-rw-r--r-- | Python/bytecodes.c | 15 | ||||
-rw-r--r-- | Python/executor_cases.c.h | 10 | ||||
-rw-r--r-- | Python/generated_cases.c.h | 13 | ||||
-rw-r--r-- | Python/specialize.c | 3 | ||||
-rw-r--r-- | Tools/cases_generator/analyzer.py | 1 |
13 files changed, 419 insertions, 142 deletions
diff --git a/Include/cpython/object.h b/Include/cpython/object.h index 2797051..a8f5782 100644 --- a/Include/cpython/object.h +++ b/Include/cpython/object.h @@ -493,6 +493,7 @@ do { \ PyAPI_FUNC(void *) PyObject_GetItemData(PyObject *obj); PyAPI_FUNC(int) PyObject_VisitManagedDict(PyObject *obj, visitproc visit, void *arg); +PyAPI_FUNC(void) _PyObject_SetManagedDict(PyObject *obj, PyObject *new_dict); PyAPI_FUNC(void) PyObject_ClearManagedDict(PyObject *obj); #define TYPE_MAX_WATCHERS 8 diff --git a/Include/internal/pycore_dict.h b/Include/internal/pycore_dict.h index fba0dfc..f33026d 100644 --- a/Include/internal/pycore_dict.h +++ b/Include/internal/pycore_dict.h @@ -1,4 +1,3 @@ - #ifndef Py_INTERNAL_DICT_H #define Py_INTERNAL_DICT_H #ifdef __cplusplus @@ -9,9 +8,10 @@ extern "C" { # error "this header requires Py_BUILD_CORE define" #endif -#include "pycore_freelist.h" // _PyFreeListState -#include "pycore_identifier.h" // _Py_Identifier -#include "pycore_object.h" // PyManagedDictPointer +#include "pycore_freelist.h" // _PyFreeListState +#include "pycore_identifier.h" // _Py_Identifier +#include "pycore_object.h" // PyManagedDictPointer +#include "pycore_pyatomic_ft_wrappers.h" // FT_ATOMIC_LOAD_SSIZE_ACQUIRE // Unsafe flavor of PyDict_GetItemWithError(): no error checking extern PyObject* _PyDict_GetItemWithError(PyObject *dp, PyObject *key); @@ -249,7 +249,7 @@ _PyDict_NotifyEvent(PyInterpreterState *interp, return DICT_NEXT_VERSION(interp) | (mp->ma_version_tag & DICT_WATCHER_AND_MODIFICATION_MASK); } -extern PyDictObject *_PyObject_MakeDictFromInstanceAttributes(PyObject *obj); +extern PyDictObject *_PyObject_MaterializeManagedDict(PyObject *obj); PyAPI_FUNC(PyObject *)_PyDict_FromItems( PyObject *const *keys, Py_ssize_t keys_offset, @@ -277,7 +277,6 @@ _PyDictValues_AddToInsertionOrder(PyDictValues *values, Py_ssize_t ix) static inline size_t shared_keys_usable_size(PyDictKeysObject *keys) { -#ifdef Py_GIL_DISABLED // dk_usable will decrease for each instance that is created and each // value that is added. dk_nentries will increase for each value that // is added. We want to always return the right value or larger. @@ -285,11 +284,9 @@ shared_keys_usable_size(PyDictKeysObject *keys) // second, and conversely here we read dk_usable first and dk_entries // second (to avoid the case where we read entries before the increment // and read usable after the decrement) - return (size_t)(_Py_atomic_load_ssize_acquire(&keys->dk_usable) + - _Py_atomic_load_ssize_acquire(&keys->dk_nentries)); -#else - return (size_t)keys->dk_nentries + (size_t)keys->dk_usable; -#endif + Py_ssize_t dk_usable = FT_ATOMIC_LOAD_SSIZE_ACQUIRE(keys->dk_usable); + Py_ssize_t dk_nentries = FT_ATOMIC_LOAD_SSIZE_ACQUIRE(keys->dk_nentries); + return dk_nentries + dk_usable; } static inline size_t diff --git a/Include/internal/pycore_object.h b/Include/internal/pycore_object.h index 512f7a3..88b052f 100644 --- a/Include/internal/pycore_object.h +++ b/Include/internal/pycore_object.h @@ -12,6 +12,7 @@ extern "C" { #include "pycore_gc.h" // _PyObject_GC_IS_TRACKED() #include "pycore_emscripten_trampoline.h" // _PyCFunction_TrampolineCall() #include "pycore_interp.h" // PyInterpreterState.gc +#include "pycore_pyatomic_ft_wrappers.h" // FT_ATOMIC_STORE_PTR_RELAXED #include "pycore_pystate.h" // _PyInterpreterState_GET() /* Check if an object is consistent. For example, ensure that the reference @@ -659,10 +660,10 @@ extern PyObject* _PyType_GetDocFromInternalDoc(const char *, const char *); extern PyObject* _PyType_GetTextSignatureFromInternalDoc(const char *, const char *, int); void _PyObject_InitInlineValues(PyObject *obj, PyTypeObject *tp); -extern int _PyObject_StoreInstanceAttribute(PyObject *obj, PyDictValues *values, - PyObject *name, PyObject *value); -PyObject * _PyObject_GetInstanceAttribute(PyObject *obj, PyDictValues *values, - PyObject *name); +extern int _PyObject_StoreInstanceAttribute(PyObject *obj, + PyObject *name, PyObject *value); +extern bool _PyObject_TryGetInstanceAttribute(PyObject *obj, PyObject *name, + PyObject **attr); #ifdef Py_GIL_DISABLED # define MANAGED_DICT_OFFSET (((Py_ssize_t)sizeof(PyObject *))*-1) @@ -683,6 +684,13 @@ _PyObject_ManagedDictPointer(PyObject *obj) return (PyManagedDictPointer *)((char *)obj + MANAGED_DICT_OFFSET); } +static inline PyDictObject * +_PyObject_GetManagedDict(PyObject *obj) +{ + PyManagedDictPointer *dorv = _PyObject_ManagedDictPointer(obj); + return (PyDictObject *)FT_ATOMIC_LOAD_PTR_RELAXED(dorv->dict); +} + static inline PyDictValues * _PyObject_InlineValues(PyObject *obj) { diff --git a/Include/internal/pycore_pyatomic_ft_wrappers.h b/Include/internal/pycore_pyatomic_ft_wrappers.h index fed5d6e..bbfc462 100644 --- a/Include/internal/pycore_pyatomic_ft_wrappers.h +++ b/Include/internal/pycore_pyatomic_ft_wrappers.h @@ -21,7 +21,10 @@ extern "C" { #ifdef Py_GIL_DISABLED #define FT_ATOMIC_LOAD_PTR(value) _Py_atomic_load_ptr(&value) +#define FT_ATOMIC_STORE_PTR(value, new_value) _Py_atomic_store_ptr(&value, new_value) #define FT_ATOMIC_LOAD_SSIZE(value) _Py_atomic_load_ssize(&value) +#define FT_ATOMIC_LOAD_SSIZE_ACQUIRE(value) \ + _Py_atomic_load_ssize_acquire(&value) #define FT_ATOMIC_LOAD_SSIZE_RELAXED(value) \ _Py_atomic_load_ssize_relaxed(&value) #define FT_ATOMIC_STORE_PTR(value, new_value) \ @@ -30,6 +33,12 @@ extern "C" { _Py_atomic_load_ptr_acquire(&value) #define FT_ATOMIC_LOAD_UINTPTR_ACQUIRE(value) \ _Py_atomic_load_uintptr_acquire(&value) +#define FT_ATOMIC_LOAD_PTR_RELAXED(value) \ + _Py_atomic_load_ptr_relaxed(&value) +#define FT_ATOMIC_LOAD_UINT8(value) \ + _Py_atomic_load_uint8(&value) +#define FT_ATOMIC_STORE_UINT8(value, new_value) \ + _Py_atomic_store_uint8(&value, new_value) #define FT_ATOMIC_STORE_PTR_RELAXED(value, new_value) \ _Py_atomic_store_ptr_relaxed(&value, new_value) #define FT_ATOMIC_STORE_PTR_RELEASE(value, new_value) \ @@ -43,11 +52,16 @@ extern "C" { #else #define FT_ATOMIC_LOAD_PTR(value) value +#define FT_ATOMIC_STORE_PTR(value, new_value) value = new_value #define FT_ATOMIC_LOAD_SSIZE(value) value +#define FT_ATOMIC_LOAD_SSIZE_ACQUIRE(value) value #define FT_ATOMIC_LOAD_SSIZE_RELAXED(value) value #define FT_ATOMIC_STORE_PTR(value, new_value) value = new_value #define FT_ATOMIC_LOAD_PTR_ACQUIRE(value) value #define FT_ATOMIC_LOAD_UINTPTR_ACQUIRE(value) value +#define FT_ATOMIC_LOAD_PTR_RELAXED(value) value +#define FT_ATOMIC_LOAD_UINT8(value) value +#define FT_ATOMIC_STORE_UINT8(value, new_value) value = new_value #define FT_ATOMIC_STORE_PTR_RELAXED(value, new_value) value = new_value #define FT_ATOMIC_STORE_PTR_RELEASE(value, new_value) value = new_value #define FT_ATOMIC_STORE_UINTPTR_RELEASE(value, new_value) value = new_value diff --git a/Lib/test/test_class.py b/Lib/test/test_class.py index a9cfd8d..5885db8 100644 --- a/Lib/test/test_class.py +++ b/Lib/test/test_class.py @@ -873,6 +873,15 @@ class TestInlineValues(unittest.TestCase): obj.foo = None # Aborted here self.assertEqual(obj.__dict__, {"foo":None}) + def test_store_attr_deleted_dict(self): + class Foo: + pass + + f = Foo() + del f.__dict__ + f.a = 3 + self.assertEqual(f.a, 3) + if __name__ == '__main__': unittest.main() diff --git a/Objects/dictobject.c b/Objects/dictobject.c index 4e69641..2644516 100644 --- a/Objects/dictobject.c +++ b/Objects/dictobject.c @@ -1752,7 +1752,7 @@ insertdict(PyInterpreterState *interp, PyDictObject *mp, uint64_t new_version = _PyDict_NotifyEvent( interp, PyDict_EVENT_MODIFIED, mp, key, value); if (_PyDict_HasSplitTable(mp)) { - mp->ma_values->values[ix] = value; + STORE_SPLIT_VALUE(mp, ix, value); if (old_value == NULL) { _PyDictValues_AddToInsertionOrder(mp->ma_values, ix); mp->ma_used++; @@ -2514,7 +2514,7 @@ delitem_common(PyDictObject *mp, Py_hash_t hash, Py_ssize_t ix, mp->ma_version_tag = new_version; if (_PyDict_HasSplitTable(mp)) { assert(old_value == mp->ma_values->values[ix]); - mp->ma_values->values[ix] = NULL; + STORE_SPLIT_VALUE(mp, ix, NULL); assert(ix < SHARED_KEYS_MAX_SIZE); /* Update order */ delete_index_from_values(mp->ma_values, ix); @@ -4226,7 +4226,7 @@ dict_setdefault_ref_lock_held(PyObject *d, PyObject *key, PyObject *default_valu assert(_PyDict_HasSplitTable(mp)); assert(mp->ma_values->values[ix] == NULL); MAINTAIN_TRACKING(mp, key, value); - mp->ma_values->values[ix] = Py_NewRef(value); + STORE_SPLIT_VALUE(mp, ix, Py_NewRef(value)); _PyDictValues_AddToInsertionOrder(mp->ma_values, ix); mp->ma_used++; mp->ma_version_tag = new_version; @@ -6616,28 +6616,79 @@ make_dict_from_instance_attributes(PyInterpreterState *interp, return res; } -PyDictObject * -_PyObject_MakeDictFromInstanceAttributes(PyObject *obj) +static PyDictObject * +materialize_managed_dict_lock_held(PyObject *obj) { + _Py_CRITICAL_SECTION_ASSERT_OBJECT_LOCKED(obj); + + OBJECT_STAT_INC(dict_materialized_on_request); + PyDictValues *values = _PyObject_InlineValues(obj); PyInterpreterState *interp = _PyInterpreterState_GET(); PyDictKeysObject *keys = CACHED_KEYS(Py_TYPE(obj)); OBJECT_STAT_INC(dict_materialized_on_request); - return make_dict_from_instance_attributes(interp, keys, values); + PyDictObject *dict = make_dict_from_instance_attributes(interp, keys, values); + FT_ATOMIC_STORE_PTR_RELEASE(_PyObject_ManagedDictPointer(obj)->dict, + (PyDictObject *)dict); + return dict; } +PyDictObject * +_PyObject_MaterializeManagedDict(PyObject *obj) +{ + PyDictObject *dict = _PyObject_GetManagedDict(obj); + if (dict != NULL) { + return dict; + } + Py_BEGIN_CRITICAL_SECTION(obj); -int -_PyObject_StoreInstanceAttribute(PyObject *obj, PyDictValues *values, +#ifdef Py_GIL_DISABLED + dict = _PyObject_GetManagedDict(obj); + if (dict != NULL) { + // We raced with another thread creating the dict + goto exit; + } +#endif + dict = materialize_managed_dict_lock_held(obj); + +#ifdef Py_GIL_DISABLED +exit: +#endif + Py_END_CRITICAL_SECTION(); + return dict; +} + +static int +set_or_del_lock_held(PyDictObject *dict, PyObject *name, PyObject *value) +{ + if (value == NULL) { + Py_hash_t hash; + if (!PyUnicode_CheckExact(name) || (hash = unicode_get_hash(name)) == -1) { + hash = PyObject_Hash(name); + if (hash == -1) + return -1; + } + return delitem_knownhash_lock_held((PyObject *)dict, name, hash); + } else { + return setitem_lock_held(dict, name, value); + } +} + +// Called with either the object's lock or the dict's lock held +// depending on whether or not a dict has been materialized for +// the object. +static int +store_instance_attr_lock_held(PyObject *obj, PyDictValues *values, PyObject *name, PyObject *value) { - PyInterpreterState *interp = _PyInterpreterState_GET(); PyDictKeysObject *keys = CACHED_KEYS(Py_TYPE(obj)); assert(keys != NULL); assert(values != NULL); assert(Py_TYPE(obj)->tp_flags & Py_TPFLAGS_INLINE_VALUES); Py_ssize_t ix = DKIX_EMPTY; + PyDictObject *dict = _PyObject_GetManagedDict(obj); + assert(dict == NULL || ((PyDictObject *)dict)->ma_values == values); if (PyUnicode_CheckExact(name)) { Py_hash_t hash = unicode_get_hash(name); if (hash == -1) { @@ -6674,25 +6725,33 @@ _PyObject_StoreInstanceAttribute(PyObject *obj, PyDictValues *values, } #endif } - PyDictObject *dict = _PyObject_ManagedDictPointer(obj)->dict; + if (ix == DKIX_EMPTY) { + int res; if (dict == NULL) { - dict = make_dict_from_instance_attributes( - interp, keys, values); - if (dict == NULL) { + // Make the dict but don't publish it in the object + // so that no one else will see it. + dict = make_dict_from_instance_attributes(PyInterpreterState_Get(), keys, values); + if (dict == NULL || + set_or_del_lock_held(dict, name, value) < 0) { + Py_XDECREF(dict); return -1; } - _PyObject_ManagedDictPointer(obj)->dict = (PyDictObject *)dict; - } - if (value == NULL) { - return PyDict_DelItem((PyObject *)dict, name); - } - else { - return PyDict_SetItem((PyObject *)dict, name, value); + + FT_ATOMIC_STORE_PTR_RELEASE(_PyObject_ManagedDictPointer(obj)->dict, + (PyDictObject *)dict); + return 0; } + + _Py_CRITICAL_SECTION_ASSERT_OBJECT_LOCKED(dict); + + res = set_or_del_lock_held (dict, name, value); + return res; } + PyObject *old_value = values->values[ix]; - values->values[ix] = Py_XNewRef(value); + FT_ATOMIC_STORE_PTR_RELEASE(values->values[ix], Py_XNewRef(value)); + if (old_value == NULL) { if (value == NULL) { PyErr_Format(PyExc_AttributeError, @@ -6719,6 +6778,72 @@ _PyObject_StoreInstanceAttribute(PyObject *obj, PyDictValues *values, return 0; } +static inline int +store_instance_attr_dict(PyObject *obj, PyDictObject *dict, PyObject *name, PyObject *value) +{ + PyDictValues *values = _PyObject_InlineValues(obj); + int res; + Py_BEGIN_CRITICAL_SECTION(dict); + if (dict->ma_values == values) { + res = store_instance_attr_lock_held(obj, values, name, value); + } + else { + res = set_or_del_lock_held(dict, name, value); + } + Py_END_CRITICAL_SECTION(); + return res; +} + +int +_PyObject_StoreInstanceAttribute(PyObject *obj, PyObject *name, PyObject *value) +{ + PyDictValues *values = _PyObject_InlineValues(obj); + if (!FT_ATOMIC_LOAD_UINT8(values->valid)) { + PyDictObject *dict = _PyObject_GetManagedDict(obj); + if (dict == NULL) { + dict = (PyDictObject *)PyObject_GenericGetDict(obj, NULL); + if (dict == NULL) { + return -1; + } + int res = store_instance_attr_dict(obj, dict, name, value); + Py_DECREF(dict); + return res; + } + return store_instance_attr_dict(obj, dict, name, value); + } + +#ifdef Py_GIL_DISABLED + // We have a valid inline values, at least for now... There are two potential + // races with having the values become invalid. One is the dictionary + // being detached from the object. The other is if someone is inserting + // into the dictionary directly and therefore causing it to resize. + // + // If we haven't materialized the dictionary yet we lock on the object, which + // will also be used to prevent the dictionary from being materialized while + // we're doing the insertion. If we race and the dictionary gets created + // then we'll need to release the object lock and lock the dictionary to + // prevent resizing. + PyDictObject *dict = _PyObject_GetManagedDict(obj); + if (dict == NULL) { + int res; + Py_BEGIN_CRITICAL_SECTION(obj); + dict = _PyObject_GetManagedDict(obj); + + if (dict == NULL) { + res = store_instance_attr_lock_held(obj, values, name, value); + } + Py_END_CRITICAL_SECTION(); + + if (dict == NULL) { + return res; + } + } + return store_instance_attr_dict(obj, dict, name, value); +#else + return store_instance_attr_lock_held(obj, values, name, value); +#endif +} + /* Sanity check for managed dicts */ #if 0 #define CHECK(val) assert(val); if (!(val)) { return 0; } @@ -6750,19 +6875,79 @@ _PyObject_ManagedDictValidityCheck(PyObject *obj) } #endif -PyObject * -_PyObject_GetInstanceAttribute(PyObject *obj, PyDictValues *values, - PyObject *name) +// Attempts to get an instance attribute from the inline values. Returns true +// if successful, or false if the caller needs to lookup in the dictionary. +bool +_PyObject_TryGetInstanceAttribute(PyObject *obj, PyObject *name, PyObject **attr) { assert(PyUnicode_CheckExact(name)); + PyDictValues *values = _PyObject_InlineValues(obj); + if (!FT_ATOMIC_LOAD_UINT8(values->valid)) { + return false; + } + PyDictKeysObject *keys = CACHED_KEYS(Py_TYPE(obj)); assert(keys != NULL); Py_ssize_t ix = _PyDictKeys_StringLookup(keys, name); if (ix == DKIX_EMPTY) { - return NULL; + *attr = NULL; + return true; + } + +#ifdef Py_GIL_DISABLED + PyObject *value = _Py_atomic_load_ptr_relaxed(&values->values[ix]); + if (value == NULL || _Py_TryIncrefCompare(&values->values[ix], value)) { + *attr = value; + return true; + } + + PyDictObject *dict = _PyObject_GetManagedDict(obj); + if (dict == NULL) { + // No dict, lock the object to prevent one from being + // materialized... + bool success = false; + Py_BEGIN_CRITICAL_SECTION(obj); + + dict = _PyObject_GetManagedDict(obj); + if (dict == NULL) { + // Still no dict, we can read from the values + assert(values->valid); + value = values->values[ix]; + *attr = Py_XNewRef(value); + success = true; + } + + Py_END_CRITICAL_SECTION(); + + if (success) { + return true; + } } + + // We have a dictionary, we'll need to lock it to prevent + // the values from being resized. + assert(dict != NULL); + + bool success; + Py_BEGIN_CRITICAL_SECTION(dict); + + if (dict->ma_values == values && FT_ATOMIC_LOAD_UINT8(values->valid)) { + value = _Py_atomic_load_ptr_relaxed(&values->values[ix]); + *attr = Py_XNewRef(value); + success = true; + } else { + // Caller needs to lookup from the dictionary + success = false; + } + + Py_END_CRITICAL_SECTION(); + + return success; +#else PyObject *value = values->values[ix]; - return Py_XNewRef(value); + *attr = Py_XNewRef(value); + return true; +#endif } int @@ -6775,20 +6960,19 @@ _PyObject_IsInstanceDictEmpty(PyObject *obj) PyDictObject *dict; if (tp->tp_flags & Py_TPFLAGS_INLINE_VALUES) { PyDictValues *values = _PyObject_InlineValues(obj); - if (values->valid) { + if (FT_ATOMIC_LOAD_UINT8(values->valid)) { PyDictKeysObject *keys = CACHED_KEYS(tp); for (Py_ssize_t i = 0; i < keys->dk_nentries; i++) { - if (values->values[i] != NULL) { + if (FT_ATOMIC_LOAD_PTR_RELAXED(values->values[i]) != NULL) { return 0; } } return 1; } - dict = _PyObject_ManagedDictPointer(obj)->dict; + dict = _PyObject_GetManagedDict(obj); } else if (tp->tp_flags & Py_TPFLAGS_MANAGED_DICT) { - PyManagedDictPointer* managed_dict = _PyObject_ManagedDictPointer(obj); - dict = managed_dict->dict; + dict = _PyObject_GetManagedDict(obj); } else { PyObject **dictptr = _PyObject_ComputedDictPointer(obj); @@ -6820,53 +7004,115 @@ PyObject_VisitManagedDict(PyObject *obj, visitproc visit, void *arg) return 0; } +static void +set_dict_inline_values(PyObject *obj, PyDictObject *new_dict) +{ + _Py_CRITICAL_SECTION_ASSERT_OBJECT_LOCKED(obj); + + PyDictValues *values = _PyObject_InlineValues(obj); + + Py_XINCREF(new_dict); + FT_ATOMIC_STORE_PTR(_PyObject_ManagedDictPointer(obj)->dict, new_dict); + + if (values->valid) { + FT_ATOMIC_STORE_UINT8(values->valid, 0); + for (Py_ssize_t i = 0; i < values->capacity; i++) { + Py_CLEAR(values->values[i]); + } + } +} + void -PyObject_ClearManagedDict(PyObject *obj) +_PyObject_SetManagedDict(PyObject *obj, PyObject *new_dict) { assert(Py_TYPE(obj)->tp_flags & Py_TPFLAGS_MANAGED_DICT); assert(_PyObject_InlineValuesConsistencyCheck(obj)); PyTypeObject *tp = Py_TYPE(obj); if (tp->tp_flags & Py_TPFLAGS_INLINE_VALUES) { - PyDictObject *dict = _PyObject_ManagedDictPointer(obj)->dict; - if (dict) { - _PyDict_DetachFromObject(dict, obj); - _PyObject_ManagedDictPointer(obj)->dict = NULL; - Py_DECREF(dict); - } - else { - PyDictValues *values = _PyObject_InlineValues(obj); - if (values->valid) { - for (Py_ssize_t i = 0; i < values->capacity; i++) { - Py_CLEAR(values->values[i]); - } - values->valid = 0; + PyDictObject *dict = _PyObject_GetManagedDict(obj); + if (dict == NULL) { +#ifdef Py_GIL_DISABLED + Py_BEGIN_CRITICAL_SECTION(obj); + + dict = _PyObject_ManagedDictPointer(obj)->dict; + if (dict == NULL) { + set_dict_inline_values(obj, (PyDictObject *)new_dict); + } + + Py_END_CRITICAL_SECTION(); + + if (dict == NULL) { + return; } +#else + set_dict_inline_values(obj, (PyDictObject *)new_dict); + return; +#endif } + + Py_BEGIN_CRITICAL_SECTION2(dict, obj); + + // We've locked dict, but the actual dict could have changed + // since we locked it. + dict = _PyObject_ManagedDictPointer(obj)->dict; + + FT_ATOMIC_STORE_PTR(_PyObject_ManagedDictPointer(obj)->dict, + (PyDictObject *)Py_XNewRef(new_dict)); + + _PyDict_DetachFromObject(dict, obj); + + Py_END_CRITICAL_SECTION2(); + + Py_XDECREF(dict); } else { - Py_CLEAR(_PyObject_ManagedDictPointer(obj)->dict); + PyDictObject *dict; + + Py_BEGIN_CRITICAL_SECTION(obj); + + dict = _PyObject_ManagedDictPointer(obj)->dict; + + FT_ATOMIC_STORE_PTR(_PyObject_ManagedDictPointer(obj)->dict, + (PyDictObject *)Py_XNewRef(new_dict)); + + Py_END_CRITICAL_SECTION(); + + Py_XDECREF(dict); } assert(_PyObject_InlineValuesConsistencyCheck(obj)); } +void +PyObject_ClearManagedDict(PyObject *obj) +{ + _PyObject_SetManagedDict(obj, NULL); +} + int _PyDict_DetachFromObject(PyDictObject *mp, PyObject *obj) { - assert(_PyObject_ManagedDictPointer(obj)->dict == mp); - assert(_PyObject_InlineValuesConsistencyCheck(obj)); - if (mp->ma_values == NULL || mp->ma_values != _PyObject_InlineValues(obj)) { + _Py_CRITICAL_SECTION_ASSERT_OBJECT_LOCKED(obj); + + if (FT_ATOMIC_LOAD_PTR_RELAXED(mp->ma_values) != _PyObject_InlineValues(obj)) { return 0; } + + // We could be called with an unlocked dict when the caller knows the + // values are already detached, so we assert after inline values check. + _Py_CRITICAL_SECTION_ASSERT_OBJECT_LOCKED(mp); assert(mp->ma_values->embedded == 1); assert(mp->ma_values->valid == 1); assert(Py_TYPE(obj)->tp_flags & Py_TPFLAGS_INLINE_VALUES); - Py_BEGIN_CRITICAL_SECTION(mp); - mp->ma_values = copy_values(mp->ma_values); - _PyObject_InlineValues(obj)->valid = 0; - Py_END_CRITICAL_SECTION(); - if (mp->ma_values == NULL) { + + PyDictValues *values = copy_values(mp->ma_values); + + if (values == NULL) { return -1; } + mp->ma_values = values; + + FT_ATOMIC_STORE_UINT8(_PyObject_InlineValues(obj)->valid, 0); + assert(_PyObject_InlineValuesConsistencyCheck(obj)); ASSERT_CONSISTENT(mp); return 0; @@ -6877,29 +7123,28 @@ PyObject_GenericGetDict(PyObject *obj, void *context) { PyInterpreterState *interp = _PyInterpreterState_GET(); PyTypeObject *tp = Py_TYPE(obj); + PyDictObject *dict; if (_PyType_HasFeature(tp, Py_TPFLAGS_MANAGED_DICT)) { - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(obj); - PyDictObject *dict = managed_dict->dict; + dict = _PyObject_GetManagedDict(obj); if (dict == NULL && (tp->tp_flags & Py_TPFLAGS_INLINE_VALUES) && - _PyObject_InlineValues(obj)->valid - ) { - PyDictValues *values = _PyObject_InlineValues(obj); - OBJECT_STAT_INC(dict_materialized_on_request); - dict = make_dict_from_instance_attributes( - interp, CACHED_KEYS(tp), values); - if (dict != NULL) { - managed_dict->dict = (PyDictObject *)dict; - } + FT_ATOMIC_LOAD_UINT8(_PyObject_InlineValues(obj)->valid)) { + dict = _PyObject_MaterializeManagedDict(obj); } - else { - dict = managed_dict->dict; + else if (dict == NULL) { + Py_BEGIN_CRITICAL_SECTION(obj); + + // Check again that we're not racing with someone else creating the dict + dict = _PyObject_GetManagedDict(obj); if (dict == NULL) { - dictkeys_incref(CACHED_KEYS(tp)); OBJECT_STAT_INC(dict_materialized_on_request); + dictkeys_incref(CACHED_KEYS(tp)); dict = (PyDictObject *)new_dict_with_shared_keys(interp, CACHED_KEYS(tp)); - managed_dict->dict = (PyDictObject *)dict; + FT_ATOMIC_STORE_PTR_RELEASE(_PyObject_ManagedDictPointer(obj)->dict, + (PyDictObject *)dict); } + + Py_END_CRITICAL_SECTION(); } return Py_XNewRef((PyObject *)dict); } @@ -7109,7 +7354,7 @@ _PyObject_InlineValuesConsistencyCheck(PyObject *obj) return 1; } assert(Py_TYPE(obj)->tp_flags & Py_TPFLAGS_MANAGED_DICT); - PyDictObject *dict = (PyDictObject *)_PyObject_ManagedDictPointer(obj)->dict; + PyDictObject *dict = _PyObject_GetManagedDict(obj); if (dict == NULL) { return 1; } diff --git a/Objects/object.c b/Objects/object.c index 73a1927..91bb011 100644 --- a/Objects/object.c +++ b/Objects/object.c @@ -6,6 +6,7 @@ #include "pycore_call.h" // _PyObject_CallNoArgs() #include "pycore_ceval.h" // _Py_EnterRecursiveCallTstate() #include "pycore_context.h" // _PyContextTokenMissing_Type +#include "pycore_critical_section.h" // Py_BEGIN_CRITICAL_SECTION, Py_END_CRITICAL_SECTION #include "pycore_descrobject.h" // _PyMethodWrapper_Type #include "pycore_dict.h" // _PyObject_MakeDictFromInstanceAttributes() #include "pycore_floatobject.h" // _PyFloat_DebugMallocStats() @@ -25,6 +26,7 @@ #include "pycore_typevarobject.h" // _PyTypeAlias_Type, _Py_initialize_generic #include "pycore_unionobject.h" // _PyUnion_Type + #ifdef Py_LIMITED_API // Prevent recursive call _Py_IncRef() <=> Py_INCREF() # error "Py_LIMITED_API macro must not be defined" @@ -1403,16 +1405,15 @@ _PyObject_GetDictPtr(PyObject *obj) if ((Py_TYPE(obj)->tp_flags & Py_TPFLAGS_MANAGED_DICT) == 0) { return _PyObject_ComputedDictPointer(obj); } - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(obj); - if (managed_dict->dict == NULL && Py_TYPE(obj)->tp_flags & Py_TPFLAGS_INLINE_VALUES) { - PyDictObject *dict = (PyDictObject *)_PyObject_MakeDictFromInstanceAttributes(obj); + PyDictObject *dict = _PyObject_GetManagedDict(obj); + if (dict == NULL && Py_TYPE(obj)->tp_flags & Py_TPFLAGS_INLINE_VALUES) { + dict = _PyObject_MaterializeManagedDict(obj); if (dict == NULL) { PyErr_Clear(); return NULL; } - managed_dict->dict = dict; } - return (PyObject **)&managed_dict->dict; + return (PyObject **)&_PyObject_ManagedDictPointer(obj)->dict; } PyObject * @@ -1480,10 +1481,9 @@ _PyObject_GetMethod(PyObject *obj, PyObject *name, PyObject **method) } } } - PyObject *dict; - if ((tp->tp_flags & Py_TPFLAGS_INLINE_VALUES) && _PyObject_InlineValues(obj)->valid) { - PyDictValues *values = _PyObject_InlineValues(obj); - PyObject *attr = _PyObject_GetInstanceAttribute(obj, values, name); + PyObject *dict, *attr; + if ((tp->tp_flags & Py_TPFLAGS_INLINE_VALUES) && + _PyObject_TryGetInstanceAttribute(obj, name, &attr)) { if (attr != NULL) { *method = attr; Py_XDECREF(descr); @@ -1492,8 +1492,7 @@ _PyObject_GetMethod(PyObject *obj, PyObject *name, PyObject **method) dict = NULL; } else if ((tp->tp_flags & Py_TPFLAGS_MANAGED_DICT)) { - PyManagedDictPointer* managed_dict = _PyObject_ManagedDictPointer(obj); - dict = (PyObject *)managed_dict->dict; + dict = (PyObject *)_PyObject_GetManagedDict(obj); } else { PyObject **dictptr = _PyObject_ComputedDictPointer(obj); @@ -1586,26 +1585,23 @@ _PyObject_GenericGetAttrWithDict(PyObject *obj, PyObject *name, } } if (dict == NULL) { - if ((tp->tp_flags & Py_TPFLAGS_INLINE_VALUES) && _PyObject_InlineValues(obj)->valid) { - PyDictValues *values = _PyObject_InlineValues(obj); - if (PyUnicode_CheckExact(name)) { - res = _PyObject_GetInstanceAttribute(obj, values, name); + if ((tp->tp_flags & Py_TPFLAGS_INLINE_VALUES)) { + if (PyUnicode_CheckExact(name) && + _PyObject_TryGetInstanceAttribute(obj, name, &res)) { if (res != NULL) { goto done; } } else { - dict = (PyObject *)_PyObject_MakeDictFromInstanceAttributes(obj); + dict = (PyObject *)_PyObject_MaterializeManagedDict(obj); if (dict == NULL) { res = NULL; goto done; } - _PyObject_ManagedDictPointer(obj)->dict = (PyDictObject *)dict; } } else if ((tp->tp_flags & Py_TPFLAGS_MANAGED_DICT)) { - PyManagedDictPointer* managed_dict = _PyObject_ManagedDictPointer(obj); - dict = (PyObject *)managed_dict->dict; + dict = (PyObject *)_PyObject_GetManagedDict(obj); } else { PyObject **dictptr = _PyObject_ComputedDictPointer(obj); @@ -1700,12 +1696,13 @@ _PyObject_GenericSetAttrWithDict(PyObject *obj, PyObject *name, if (dict == NULL) { PyObject **dictptr; - if ((tp->tp_flags & Py_TPFLAGS_INLINE_VALUES) && _PyObject_InlineValues(obj)->valid) { - res = _PyObject_StoreInstanceAttribute( - obj, _PyObject_InlineValues(obj), name, value); + + if ((tp->tp_flags & Py_TPFLAGS_INLINE_VALUES)) { + res = _PyObject_StoreInstanceAttribute(obj, name, value); goto error_check; } - else if ((tp->tp_flags & Py_TPFLAGS_MANAGED_DICT)) { + + if ((tp->tp_flags & Py_TPFLAGS_MANAGED_DICT)) { PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(obj); dictptr = (PyObject **)&managed_dict->dict; } @@ -1779,7 +1776,7 @@ PyObject_GenericSetDict(PyObject *obj, PyObject *value, void *context) PyObject **dictptr = _PyObject_GetDictPtr(obj); if (dictptr == NULL) { if (_PyType_HasFeature(Py_TYPE(obj), Py_TPFLAGS_INLINE_VALUES) && - _PyObject_ManagedDictPointer(obj)->dict == NULL + _PyObject_GetManagedDict(obj) == NULL ) { /* Was unable to convert to dict */ PyErr_NoMemory(); diff --git a/Objects/typeobject.c b/Objects/typeobject.c index 970c82d..808e11f 100644 --- a/Objects/typeobject.c +++ b/Objects/typeobject.c @@ -3165,9 +3165,9 @@ subtype_setdict(PyObject *obj, PyObject *value, void *context) "not a '%.200s'", Py_TYPE(value)->tp_name); return -1; } + if (Py_TYPE(obj)->tp_flags & Py_TPFLAGS_MANAGED_DICT) { - PyObject_ClearManagedDict(obj); - _PyObject_ManagedDictPointer(obj)->dict = (PyDictObject *)Py_XNewRef(value); + _PyObject_SetManagedDict(obj, value); } else { dictptr = _PyObject_ComputedDictPointer(obj); @@ -6194,15 +6194,27 @@ object_set_class(PyObject *self, PyObject *value, void *closure) /* Changing the class will change the implicit dict keys, * so we must materialize the dictionary first. */ if (oldto->tp_flags & Py_TPFLAGS_INLINE_VALUES) { - PyDictObject *dict = _PyObject_ManagedDictPointer(self)->dict; + PyDictObject *dict = _PyObject_MaterializeManagedDict(self); if (dict == NULL) { - dict = (PyDictObject *)_PyObject_MakeDictFromInstanceAttributes(self); - if (dict == NULL) { - return -1; - } - _PyObject_ManagedDictPointer(self)->dict = dict; + return -1; + } + + bool error = false; + + Py_BEGIN_CRITICAL_SECTION2(self, dict); + + // If we raced after materialization and replaced the dict + // then the materialized dict should no longer have the + // inline values in which case detach is a nop. + assert(_PyObject_GetManagedDict(self) == dict || + dict->ma_values != _PyObject_InlineValues(self)); + + if (_PyDict_DetachFromObject(dict, self) < 0) { + error = true; } - if (_PyDict_DetachFromObject(dict, self)) { + + Py_END_CRITICAL_SECTION2(); + if (error) { return -1; } } diff --git a/Python/bytecodes.c b/Python/bytecodes.c index c1fbd3c..b7511b9 100644 --- a/Python/bytecodes.c +++ b/Python/bytecodes.c @@ -1947,15 +1947,13 @@ dummy_func( op(_CHECK_ATTR_WITH_HINT, (owner -- owner)) { assert(Py_TYPE(owner)->tp_flags & Py_TPFLAGS_MANAGED_DICT); - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); DEOPT_IF(dict == NULL); assert(PyDict_CheckExact((PyObject *)dict)); } op(_LOAD_ATTR_WITH_HINT, (hint/1, owner -- attr, null if (oparg & 1))) { - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); DEOPT_IF(hint >= (size_t)dict->ma_keys->dk_nentries); PyObject *name = GETITEM(FRAME_CO_NAMES, oparg>>1); if (DK_IS_UNICODE(dict->ma_keys)) { @@ -2072,14 +2070,15 @@ dummy_func( op(_GUARD_DORV_NO_DICT, (owner -- owner)) { assert(Py_TYPE(owner)->tp_dictoffset < 0); assert(Py_TYPE(owner)->tp_flags & Py_TPFLAGS_INLINE_VALUES); - DEOPT_IF(_PyObject_ManagedDictPointer(owner)->dict); + DEOPT_IF(_PyObject_GetManagedDict(owner)); DEOPT_IF(_PyObject_InlineValues(owner)->valid == 0); } op(_STORE_ATTR_INSTANCE_VALUE, (index/1, value, owner --)) { STAT_INC(STORE_ATTR, hit); - assert(_PyObject_ManagedDictPointer(owner)->dict == NULL); + assert(_PyObject_GetManagedDict(owner) == NULL); PyDictValues *values = _PyObject_InlineValues(owner); + PyObject *old_value = values->values[index]; values->values[index] = value; if (old_value == NULL) { @@ -2088,6 +2087,7 @@ dummy_func( else { Py_DECREF(old_value); } + Py_DECREF(owner); } @@ -2102,8 +2102,7 @@ dummy_func( assert(type_version != 0); DEOPT_IF(tp->tp_version_tag != type_version); assert(tp->tp_flags & Py_TPFLAGS_MANAGED_DICT); - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); DEOPT_IF(dict == NULL); assert(PyDict_CheckExact((PyObject *)dict)); PyObject *name = GETITEM(FRAME_CO_NAMES, oparg); diff --git a/Python/executor_cases.c.h b/Python/executor_cases.c.h index df87f91..841ce8c 100644 --- a/Python/executor_cases.c.h +++ b/Python/executor_cases.c.h @@ -1998,8 +1998,7 @@ PyObject *owner; owner = stack_pointer[-1]; assert(Py_TYPE(owner)->tp_flags & Py_TPFLAGS_MANAGED_DICT); - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); if (dict == NULL) { UOP_STAT_INC(uopcode, miss); JUMP_TO_JUMP_TARGET(); @@ -2015,8 +2014,7 @@ oparg = CURRENT_OPARG(); owner = stack_pointer[-1]; uint16_t hint = (uint16_t)CURRENT_OPERAND(); - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); if (hint >= (size_t)dict->ma_keys->dk_nentries) { UOP_STAT_INC(uopcode, miss); JUMP_TO_JUMP_TARGET(); @@ -2159,7 +2157,7 @@ owner = stack_pointer[-1]; assert(Py_TYPE(owner)->tp_dictoffset < 0); assert(Py_TYPE(owner)->tp_flags & Py_TPFLAGS_INLINE_VALUES); - if (_PyObject_ManagedDictPointer(owner)->dict) { + if (_PyObject_GetManagedDict(owner)) { UOP_STAT_INC(uopcode, miss); JUMP_TO_JUMP_TARGET(); } @@ -2177,7 +2175,7 @@ value = stack_pointer[-2]; uint16_t index = (uint16_t)CURRENT_OPERAND(); STAT_INC(STORE_ATTR, hit); - assert(_PyObject_ManagedDictPointer(owner)->dict == NULL); + assert(_PyObject_GetManagedDict(owner) == NULL); PyDictValues *values = _PyObject_InlineValues(owner); PyObject *old_value = values->values[index]; values->values[index] = value; diff --git a/Python/generated_cases.c.h b/Python/generated_cases.c.h index a426d9e..058cac8 100644 --- a/Python/generated_cases.c.h +++ b/Python/generated_cases.c.h @@ -4017,16 +4017,14 @@ // _CHECK_ATTR_WITH_HINT { assert(Py_TYPE(owner)->tp_flags & Py_TPFLAGS_MANAGED_DICT); - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); DEOPT_IF(dict == NULL, LOAD_ATTR); assert(PyDict_CheckExact((PyObject *)dict)); } // _LOAD_ATTR_WITH_HINT { uint16_t hint = read_u16(&this_instr[4].cache); - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); DEOPT_IF(hint >= (size_t)dict->ma_keys->dk_nentries, LOAD_ATTR); PyObject *name = GETITEM(FRAME_CO_NAMES, oparg>>1); if (DK_IS_UNICODE(dict->ma_keys)) { @@ -5309,7 +5307,7 @@ { assert(Py_TYPE(owner)->tp_dictoffset < 0); assert(Py_TYPE(owner)->tp_flags & Py_TPFLAGS_INLINE_VALUES); - DEOPT_IF(_PyObject_ManagedDictPointer(owner)->dict, STORE_ATTR); + DEOPT_IF(_PyObject_GetManagedDict(owner), STORE_ATTR); DEOPT_IF(_PyObject_InlineValues(owner)->valid == 0, STORE_ATTR); } // _STORE_ATTR_INSTANCE_VALUE @@ -5317,7 +5315,7 @@ { uint16_t index = read_u16(&this_instr[4].cache); STAT_INC(STORE_ATTR, hit); - assert(_PyObject_ManagedDictPointer(owner)->dict == NULL); + assert(_PyObject_GetManagedDict(owner) == NULL); PyDictValues *values = _PyObject_InlineValues(owner); PyObject *old_value = values->values[index]; values->values[index] = value; @@ -5380,8 +5378,7 @@ assert(type_version != 0); DEOPT_IF(tp->tp_version_tag != type_version, STORE_ATTR); assert(tp->tp_flags & Py_TPFLAGS_MANAGED_DICT); - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); DEOPT_IF(dict == NULL, STORE_ATTR); assert(PyDict_CheckExact((PyObject *)dict)); PyObject *name = GETITEM(FRAME_CO_NAMES, oparg); diff --git a/Python/specialize.c b/Python/specialize.c index 5e14bb5..ee51781 100644 --- a/Python/specialize.c +++ b/Python/specialize.c @@ -852,8 +852,7 @@ specialize_dict_access( instr->op.code = values_op; } else { - PyManagedDictPointer *managed_dict = _PyObject_ManagedDictPointer(owner); - PyDictObject *dict = managed_dict->dict; + PyDictObject *dict = _PyObject_GetManagedDict(owner); if (dict == NULL || !PyDict_CheckExact(dict)) { SPECIALIZATION_FAIL(base_op, SPEC_FAIL_NO_DICT); return 0; diff --git a/Tools/cases_generator/analyzer.py b/Tools/cases_generator/analyzer.py index d17b2b9..18cefa0 100644 --- a/Tools/cases_generator/analyzer.py +++ b/Tools/cases_generator/analyzer.py @@ -354,6 +354,7 @@ def has_error_without_pop(op: parser.InstDef) -> bool: NON_ESCAPING_FUNCTIONS = ( "Py_INCREF", "_PyManagedDictPointer_IsValues", + "_PyObject_GetManagedDict", "_PyObject_ManagedDictPointer", "_PyObject_InlineValues", "_PyDictValues_AddToInsertionOrder", |