GH-101578: Normalize the current exception (GH-101607)

* Make sure that the current exception is always normalized.

* Remove redundant type and traceback fields for the current exception.

* Add new API functions: PyErr_GetRaisedException, PyErr_SetRaisedException

* Add new API functions: PyException_GetArgs, PyException_SetArgs

Python/bytecodes.c

@@ -804,9 +804,7 @@ dummy_func(
                 DECREF_INPUTS();
             }
             else {
-                PyObject *exc_type = Py_NewRef(Py_TYPE(exc_value));
-                PyObject *exc_traceback = PyException_GetTraceback(exc_value);
-                _PyErr_Restore(tstate, exc_type, Py_NewRef(exc_value), exc_traceback);
+                _PyErr_SetRaisedException(tstate, Py_NewRef(exc_value));
                 goto exception_unwind;
             }
         }

Python/ceval.c

@@ -2902,13 +2902,13 @@ format_kwargs_error(PyThreadState *tstate, PyObject *func, PyObject *kwargs)
         }
     }
     else if (_PyErr_ExceptionMatches(tstate, PyExc_KeyError)) {
-        PyObject *exc, *val, *tb;
-        _PyErr_Fetch(tstate, &exc, &val, &tb);
-        if (val && PyTuple_Check(val) && PyTuple_GET_SIZE(val) == 1) {
+        PyObject *exc = _PyErr_GetRaisedException(tstate);
+        PyObject *args = ((PyBaseExceptionObject *)exc)->args;
+        if (exc && PyTuple_Check(args) && PyTuple_GET_SIZE(args) == 1) {
             _PyErr_Clear(tstate);
             PyObject *funcstr = _PyObject_FunctionStr(func);
             if (funcstr != NULL) {
-                PyObject *key = PyTuple_GET_ITEM(val, 0);
+                PyObject *key = PyTuple_GET_ITEM(args, 0);
                 _PyErr_Format(
                     tstate, PyExc_TypeError,
                     "%U got multiple values for keyword argument '%S'",
@@ -2916,11 +2916,9 @@ format_kwargs_error(PyThreadState *tstate, PyObject *func, PyObject *kwargs)
                 Py_DECREF(funcstr);
             }
             Py_XDECREF(exc);
-            Py_XDECREF(val);
-            Py_XDECREF(tb);
         }
         else {
-            _PyErr_Restore(tstate, exc, val, tb);
+            _PyErr_SetRaisedException(tstate, exc);
         }
     }
 }

Python/errors.c

@@ -27,54 +27,12 @@ static PyObject *
 _PyErr_FormatV(PyThreadState *tstate, PyObject *exception,
                const char *format, va_list vargs);
 
-
 void
-_PyErr_Restore(PyThreadState *tstate, PyObject *type, PyObject *value,
-               PyObject *traceback)
+_PyErr_SetRaisedException(PyThreadState *tstate, PyObject *exc)
 {
-    PyObject *oldtype, *oldvalue, *oldtraceback;
-
-    if (traceback != NULL && !PyTraceBack_Check(traceback)) {
-        /* XXX Should never happen -- fatal error instead? */
-        /* Well, it could be None. */
-        Py_SETREF(traceback, NULL);
-    }
-
-    /* Save these in locals to safeguard against recursive
-       invocation through Py_XDECREF */
-    oldtype = tstate->curexc_type;
-    oldvalue = tstate->curexc_value;
-    oldtraceback = tstate->curexc_traceback;
-
-    tstate->curexc_type = type;
-    tstate->curexc_value = value;
-    tstate->curexc_traceback = traceback;
-
-    Py_XDECREF(oldtype);
-    Py_XDECREF(oldvalue);
-    Py_XDECREF(oldtraceback);
-}
-
-void
-PyErr_Restore(PyObject *type, PyObject *value, PyObject *traceback)
-{
-    PyThreadState *tstate = _PyThreadState_GET();
-    _PyErr_Restore(tstate, type, value, traceback);
-}
-
-
-_PyErr_StackItem *
-_PyErr_GetTopmostException(PyThreadState *tstate)
-{
-    _PyErr_StackItem *exc_info = tstate->exc_info;
-    assert(exc_info);
-
-    while ((exc_info->exc_value == NULL || exc_info->exc_value == Py_None) &&
-           exc_info->previous_item != NULL)
-    {
-        exc_info = exc_info->previous_item;
-    }
-    return exc_info;
+    PyObject *old_exc = tstate->current_exception;
+    tstate->current_exception = exc;
+    Py_XDECREF(old_exc);
 }
 
 static PyObject*
@@ -103,6 +61,80 @@ _PyErr_CreateException(PyObject *exception_type, PyObject *value)
     return exc;
 }
 
+void
+_PyErr_Restore(PyThreadState *tstate, PyObject *type, PyObject *value,
+               PyObject *traceback)
+{
+    if (type == NULL) {
+        assert(value == NULL);
+        assert(traceback == NULL);
+        _PyErr_SetRaisedException(tstate, NULL);
+        return;
+    }
+    assert(PyExceptionClass_Check(type));
+    if (value != NULL && type == (PyObject *)Py_TYPE(value)) {
+        /* Already normalized */
+        assert(((PyBaseExceptionObject *)value)->traceback != Py_None);
+    }
+    else {
+        PyObject *exc = _PyErr_CreateException(type, value);
+        Py_XDECREF(value);
+        if (exc == NULL) {
+            Py_DECREF(type);
+            Py_XDECREF(traceback);
+            return;
+        }
+        value = exc;
+    }
+    assert(PyExceptionInstance_Check(value));
+    if (traceback != NULL && !PyTraceBack_Check(traceback)) {
+        if (traceback == Py_None) {
+            Py_DECREF(Py_None);
+            traceback = NULL;
+        }
+        else {
+            PyErr_SetString(PyExc_TypeError, "traceback must be a Traceback or None");
+            Py_XDECREF(value);
+            Py_DECREF(type);
+            Py_XDECREF(traceback);
+            return;
+        }
+    }
+    PyObject *old_traceback = ((PyBaseExceptionObject *)value)->traceback;
+    ((PyBaseExceptionObject *)value)->traceback = traceback;
+    Py_XDECREF(old_traceback);
+    _PyErr_SetRaisedException(tstate, value);
+    Py_DECREF(type);
+}
+
+void
+PyErr_Restore(PyObject *type, PyObject *value, PyObject *traceback)
+{
+    PyThreadState *tstate = _PyThreadState_GET();
+    _PyErr_Restore(tstate, type, value, traceback);
+}
+
+void
+PyErr_SetRaisedException(PyObject *exc)
+{
+    PyThreadState *tstate = _PyThreadState_GET();
+    _PyErr_SetRaisedException(tstate, exc);
+}
+
+_PyErr_StackItem *
+_PyErr_GetTopmostException(PyThreadState *tstate)
+{
+    _PyErr_StackItem *exc_info = tstate->exc_info;
+    assert(exc_info);
+
+    while ((exc_info->exc_value == NULL || exc_info->exc_value == Py_None) &&
+           exc_info->previous_item != NULL)
+    {
+        exc_info = exc_info->previous_item;
+    }
+    return exc_info;
+}
+
 void
 _PyErr_SetObject(PyThreadState *tstate, PyObject *exception, PyObject *value)
 {
@@ -117,30 +149,29 @@ _PyErr_SetObject(PyThreadState *tstate, PyObject *exception, PyObject *value)
                       exception);
         return;
     }
-
     Py_XINCREF(value);
+    /* Normalize the exception */
+    if (value == NULL || (PyObject *)Py_TYPE(value) != exception) {
+        /* We must normalize the value right now */
+        PyObject *fixed_value;
+
+        /* Issue #23571: functions must not be called with an
+            exception set */
+        _PyErr_Clear(tstate);
+
+        fixed_value = _PyErr_CreateException(exception, value);
+        Py_XDECREF(value);
+        if (fixed_value == NULL) {
+            return;
+        }
+
+        value = fixed_value;
+    }
+
     exc_value = _PyErr_GetTopmostException(tstate)->exc_value;
     if (exc_value != NULL && exc_value != Py_None) {
         /* Implicit exception chaining */
         Py_INCREF(exc_value);
-        if (value == NULL || !PyExceptionInstance_Check(value)) {
-            /* We must normalize the value right now */
-            PyObject *fixed_value;
-
-            /* Issue #23571: functions must not be called with an
-               exception set */
-            _PyErr_Clear(tstate);
-
-            fixed_value = _PyErr_CreateException(exception, value);
-            Py_XDECREF(value);
-            if (fixed_value == NULL) {
-                Py_DECREF(exc_value);
-                return;
-            }
-
-            value = fixed_value;
-        }
-
         /* Avoid creating new reference cycles through the
            context chain, while taking care not to hang on
            pre-existing ones.
@@ -414,17 +445,34 @@ PyErr_NormalizeException(PyObject **exc, PyObject **val, PyObject **tb)
 }
 
 
+PyObject *
+_PyErr_GetRaisedException(PyThreadState *tstate) {
+    PyObject *exc = tstate->current_exception;
+    tstate->current_exception = NULL;
+    return exc;
+}
+
+PyObject *
+PyErr_GetRaisedException(void)
+{
+    PyThreadState *tstate = _PyThreadState_GET();
+    return _PyErr_GetRaisedException(tstate);
+}
+
 void
 _PyErr_Fetch(PyThreadState *tstate, PyObject **p_type, PyObject **p_value,
              PyObject **p_traceback)
 {
-    *p_type = tstate->curexc_type;
-    *p_value = tstate->curexc_value;
-    *p_traceback = tstate->curexc_traceback;
-
-    tstate->curexc_type = NULL;
-    tstate->curexc_value = NULL;
-    tstate->curexc_traceback = NULL;
+    PyObject *exc = _PyErr_GetRaisedException(tstate);
+    *p_value = exc;
+    if (exc == NULL) {
+        *p_type = NULL;
+        *p_traceback = NULL;
+    }
+    else {
+        *p_type = Py_NewRef(Py_TYPE(exc));
+        *p_traceback = Py_XNewRef(((PyBaseExceptionObject *)exc)->traceback);
+    }
 }
 
 
@@ -597,6 +645,28 @@ _PyErr_ChainExceptions(PyObject *typ, PyObject *val, PyObject *tb)
     }
 }
 
+/* Like PyErr_SetRaisedException(), but if an exception is already set,
+   set the context associated with it.
+
+   The caller is responsible for ensuring that this call won't create
+   any cycles in the exception context chain. */
+void
+_PyErr_ChainExceptions1(PyObject *exc)
+{
+    if (exc == NULL) {
+        return;
+    }
+    PyThreadState *tstate = _PyThreadState_GET();
+    if (_PyErr_Occurred(tstate)) {
+        PyObject *exc2 = _PyErr_GetRaisedException(tstate);
+        PyException_SetContext(exc2, exc);
+        _PyErr_SetRaisedException(tstate, exc2);
+    }
+    else {
+        _PyErr_SetRaisedException(tstate, exc);
+    }
+}
+
 /* Set the currently set exception's context to the given exception.
 
    If the provided exc_info is NULL, then the current Python thread state's
@@ -706,19 +776,6 @@ PyErr_BadArgument(void)
     return 0;
 }
 
-PyObject *
-_PyErr_NoMemory(PyThreadState *tstate)
-{
-    if (Py_IS_TYPE(PyExc_MemoryError, NULL)) {
-        /* PyErr_NoMemory() has been called before PyExc_MemoryError has been
-           initialized by _PyExc_Init() */
-        Py_FatalError("Out of memory and PyExc_MemoryError is not "
-                      "initialized yet");
-    }
-    _PyErr_SetNone(tstate, PyExc_MemoryError);
-    return NULL;
-}
-
 PyObject *
 PyErr_NoMemory(void)
 {

Python/generated_cases.c.h

@@ -1036,9 +1036,7 @@
                 Py_DECREF(exc_value);
             }
             else {
-                PyObject *exc_type = Py_NewRef(Py_TYPE(exc_value));
-                PyObject *exc_traceback = PyException_GetTraceback(exc_value);
-                _PyErr_Restore(tstate, exc_type, Py_NewRef(exc_value), exc_traceback);
+                _PyErr_SetRaisedException(tstate, Py_NewRef(exc_value));
                 goto exception_unwind;
             }
             STACK_SHRINK(2);

Python/import.c

@@ -1592,6 +1592,13 @@ remove_importlib_frames(PyThreadState *tstate)
         Py_DECREF(code);
         tb = next;
     }
+    assert(PyExceptionInstance_Check(value));
+    assert((PyObject *)Py_TYPE(value) == exception);
+    if (base_tb == NULL) {
+        base_tb = Py_None;
+        Py_INCREF(Py_None);
+    }
+    PyException_SetTraceback(value, base_tb);
 done:
     _PyErr_Restore(tstate, exception, value, base_tb);
 }

Python/initconfig.c

@@ -3143,8 +3143,7 @@ init_dump_ascii_wstr(const wchar_t *str)
 void
 _Py_DumpPathConfig(PyThreadState *tstate)
 {
-    PyObject *exc_type, *exc_value, *exc_tb;
-    _PyErr_Fetch(tstate, &exc_type, &exc_value, &exc_tb);
+    PyObject *exc = _PyErr_GetRaisedException(tstate);
 
     PySys_WriteStderr("Python path configuration:\n");
 
@@ -3202,5 +3201,5 @@ _Py_DumpPathConfig(PyThreadState *tstate)
         PySys_WriteStderr("  ]\n");
     }
 
-    _PyErr_Restore(tstate, exc_type, exc_value, exc_tb);
+    _PyErr_SetRaisedException(tstate, exc);
 }

Python/pystate.c

@@ -1375,9 +1375,7 @@ PyThreadState_Clear(PyThreadState *tstate)
     Py_CLEAR(tstate->dict);
     Py_CLEAR(tstate->async_exc);
 
-    Py_CLEAR(tstate->curexc_type);
-    Py_CLEAR(tstate->curexc_value);
-    Py_CLEAR(tstate->curexc_traceback);
+    Py_CLEAR(tstate->current_exception);
 
     Py_CLEAR(tstate->exc_state.exc_value);
 

Python/pythonrun.c

@@ -748,13 +748,10 @@ _Py_HandleSystemExit(int *exitcode_p)
     }
 
  done:
-    /* Restore and clear the exception info, in order to properly decref
-     * the exception, value, and traceback.      If we just exit instead,
-     * these leak, which confuses PYTHONDUMPREFS output, and may prevent
-     * some finalizers from running.
-     */
-    PyErr_Restore(exception, value, tb);
-    PyErr_Clear();
+    /* Cleanup the exception */
+    Py_CLEAR(exception);
+    Py_CLEAR(value);
+    Py_CLEAR(tb);
     *exitcode_p = exitcode;
     return 1;
 }

Python/sysmodule.c

@@ -66,12 +66,11 @@ _PySys_GetAttr(PyThreadState *tstate, PyObject *name)
     if (sd == NULL) {
         return NULL;
     }
-    PyObject *exc_type, *exc_value, *exc_tb;
-    _PyErr_Fetch(tstate, &exc_type, &exc_value, &exc_tb);
+    PyObject *exc = _PyErr_GetRaisedException(tstate);
     /* XXX Suppress a new exception if it was raised and restore
      * the old one. */
     PyObject *value = _PyDict_GetItemWithError(sd, name);
-    _PyErr_Restore(tstate, exc_type, exc_value, exc_tb);
+    _PyErr_SetRaisedException(tstate, exc);
     return value;
 }
 
@@ -3704,11 +3703,10 @@ static void
 sys_format(PyObject *key, FILE *fp, const char *format, va_list va)
 {
     PyObject *file, *message;
-    PyObject *error_type, *error_value, *error_traceback;
     const char *utf8;
     PyThreadState *tstate = _PyThreadState_GET();
 
-    _PyErr_Fetch(tstate, &error_type, &error_value, &error_traceback);
+    PyObject *error = _PyErr_GetRaisedException(tstate);
     file = _PySys_GetAttr(tstate, key);
     message = PyUnicode_FromFormatV(format, va);
     if (message != NULL) {
@@ -3720,7 +3718,7 @@ sys_format(PyObject *key, FILE *fp, const char *format, va_list va)
         }
         Py_DECREF(message);
     }
-    _PyErr_Restore(tstate, error_type, error_value, error_traceback);
+    _PyErr_SetRaisedException(tstate, error);
 }
 
 void

Python/traceback.c

@@ -249,6 +249,8 @@ PyTraceBack_Here(PyFrameObject *frame)
         _PyErr_ChainExceptions(exc, val, tb);
         return -1;
     }
+    assert(PyExceptionInstance_Check(val));
+    PyException_SetTraceback(val, newtb);
     PyErr_Restore(exc, val, newtb);
     Py_XDECREF(tb);
     return 0;
@@ -260,13 +262,12 @@ void _PyTraceback_Add(const char *funcname, const char *filename, int lineno)
     PyObject *globals;
     PyCodeObject *code;
     PyFrameObject *frame;
-    PyObject *exc, *val, *tb;
     PyThreadState *tstate = _PyThreadState_GET();
 
     /* Save and clear the current exception. Python functions must not be
        called with an exception set. Calling Python functions happens when
        the codec of the filesystem encoding is implemented in pure Python. */
-    _PyErr_Fetch(tstate, &exc, &val, &tb);
+    PyObject *exc = _PyErr_GetRaisedException(tstate);
 
     globals = PyDict_New();
     if (!globals)
@@ -283,13 +284,13 @@ void _PyTraceback_Add(const char *funcname, const char *filename, int lineno)
         goto error;
     frame->f_lineno = lineno;
 
-    _PyErr_Restore(tstate, exc, val, tb);
+    _PyErr_SetRaisedException(tstate, exc);
     PyTraceBack_Here(frame);
     Py_DECREF(frame);
     return;
 
 error:
-    _PyErr_ChainExceptions(exc, val, tb);
+    _PyErr_ChainExceptions1(exc);
 }
 
 static PyObject *