gh-133157: remove usage of `_Py_NO_SANITIZE_UNDEFINED` in `pyexpat` (#135346)

This was the last usage, so the macro is removed as well.

Co-authored-by: Bénédikt Tran <10796600+picnixz@users.noreply.github.com>

diff --git a/Include/pyport.h b/Include/pyport.h
index 0675294d5bc3b1d18996c1158cf458c195edf735..89829373be2ca2bd5a55fbda2a42c0ed5748dda6 100644 (file)
--- a/Include/pyport.h
+++ b/Include/pyport.h
@@ -667,25 +667,6 @@ extern "C" {
 #endif
 
 
-// _Py_NO_SANITIZE_UNDEFINED(): Disable Undefined Behavior sanitizer (UBsan)
-// on a function.
-//
-// Clang and GCC 9.0+ use __attribute__((no_sanitize("undefined"))).
-// GCC 4.9+ uses __attribute__((no_sanitize_undefined)).
-#if defined(__has_feature)
-#  if __has_feature(undefined_behavior_sanitizer)
-#    define _Py_NO_SANITIZE_UNDEFINED __attribute__((no_sanitize("undefined")))
-#  endif
-#endif
-#if !defined(_Py_NO_SANITIZE_UNDEFINED) && defined(__GNUC__) \
-    && ((__GNUC__ >= 5) || (__GNUC__ == 4) && (__GNUC_MINOR__ >= 9))
-#  define _Py_NO_SANITIZE_UNDEFINED __attribute__((no_sanitize_undefined))
-#endif
-#ifndef _Py_NO_SANITIZE_UNDEFINED
-#  define _Py_NO_SANITIZE_UNDEFINED
-#endif
-
-
 // _Py_NONSTRING: The nonstring variable attribute specifies that an object or
 // member declaration with type array of char, signed char, or unsigned char,
 // or pointer to such a type is intended to store character arrays that do not

diff --git a/Lib/test/test_pyexpat.py b/Lib/test/test_pyexpat.py
index 1d56ccd71cf962bcc45843e7747f3216e7d7d238..d4b4f60be980a5b48dd3e6a786b17b4a401aaa4f 100644 (file)
--- a/Lib/test/test_pyexpat.py
+++ b/Lib/test/test_pyexpat.py
@@ -9,12 +9,11 @@ import traceback
 from io import BytesIO
 from test import support
 from test.support import os_helper
-
+from test.support import sortdict
+from unittest import mock
 from xml.parsers import expat
 from xml.parsers.expat import errors
 
-from test.support import sortdict
-
 
 class SetAttributeTest(unittest.TestCase):
     def setUp(self):
@@ -436,6 +435,19 @@ class BufferTextTest(unittest.TestCase):
                           "<!--abc-->", "4", "<!--def-->", "5", "</a>"],
                          "buffered text not properly split")
 
+    def test_change_character_data_handler_in_callback(self):
+        # Test that xmlparse_handler_setter() properly handles
+        # the special case "parser.CharacterDataHandler = None".
+        def handler(*args):
+            parser.CharacterDataHandler = None
+
+        handler_wrapper = mock.Mock(wraps=handler)
+        parser = expat.ParserCreate()
+        parser.CharacterDataHandler = handler_wrapper
+        parser.Parse(b"<a>1<b/>2<c></c>3<!--abc-->4<!--def-->5</a> ", True)
+        handler_wrapper.assert_called_once()
+        self.assertIsNone(parser.CharacterDataHandler)
+
 
 # Test handling of exception from callback:
 class HandlerExceptionTest(unittest.TestCase):
@@ -595,7 +607,7 @@ class ChardataBufferTest(unittest.TestCase):
     def test_disabling_buffer(self):
         xml1 = b"<?xml version='1.0' encoding='iso8859'?><a>" + b'a' * 512
         xml2 = b'b' * 1024
-        xml3 = b'c' * 1024 + b'</a>';
+        xml3 = b'c' * 1024 + b'</a>'
         parser = expat.ParserCreate()
         parser.CharacterDataHandler = self.counting_handler
         parser.buffer_text = 1

diff --git a/Misc/NEWS.d/next/C_API/2025-06-19-12-47-18.gh-issue-133157.1WA85f.rst b/Misc/NEWS.d/next/C_API/2025-06-19-12-47-18.gh-issue-133157.1WA85f.rst
new file mode 100644 (file)
index 0000000..1b37d88
--- /dev/null
+++ b/Misc/NEWS.d/next/C_API/2025-06-19-12-47-18.gh-issue-133157.1WA85f.rst
@@ -0,0 +1 @@
+Remove the private, undocumented macro :c:macro:`!_Py_NO_SANITIZE_UNDEFINED`.

diff --git a/Modules/pyexpat.c b/Modules/pyexpat.c
index fa153d86543e9996992e22c504ca915f38c3d8a2..c449dd848d10c664fe47b95dd819438a5317363d 100644 (file)
--- a/Modules/pyexpat.c
+++ b/Modules/pyexpat.c
@@ -98,7 +98,11 @@ typedef struct {
 
 #define CHARACTER_DATA_BUFFER_SIZE 8192
 
-typedef const void *xmlhandler;
+// A generic function type for storage.
+// To avoid undefined behaviors, a handler must be cast to the correct
+// function type before it's called; see SETTER_WRAPPER below.
+typedef void (*xmlhandler)(void);
+
 typedef void (*xmlhandlersetter)(XML_Parser self, xmlhandler handler);
 
 struct HandlerInfo {
@@ -110,9 +114,7 @@ struct HandlerInfo {
 
 static struct HandlerInfo handler_info[64];
 
-// gh-111178: Use _Py_NO_SANITIZE_UNDEFINED, rather than using the exact
-// handler API for each handler.
-static inline void _Py_NO_SANITIZE_UNDEFINED
+static inline void
 CALL_XML_HANDLER_SETTER(const struct HandlerInfo *handler_info,
                         XML_Parser xml_parser, xmlhandler xml_handler)
 {
@@ -1365,7 +1367,7 @@ xmlparse_handler_setter(PyObject *op, PyObject *v, void *closure)
            elaborate system of handlers and state could remove the
            C handler more effectively. */
         if (handlernum == CharacterData && self->in_callback) {
-            c_handler = noop_character_data_handler;
+            c_handler = (xmlhandler)noop_character_data_handler;
         }
         v = NULL;
     }
@@ -2222,13 +2224,84 @@ clear_handlers(xmlparseobject *self, int initial)
     }
 }
 
+/* To avoid undefined behaviors, a function must be *called* via a function
+ * pointer of the correct type.
+ * So, for each `XML_Set*` function, we define a wrapper that calls `XML_Set*`
+ * with its argument cast to the appropriate type.
+ */
+
+typedef void (*parser_only)(void *);
+typedef int (*not_standalone)(void *);
+typedef void (*parser_and_data)(void *, const XML_Char *);
+typedef void (*parser_and_data_and_int)(void *, const XML_Char *, int);
+typedef void (*parser_and_data_and_data)(
+    void *, const XML_Char *, const XML_Char *);
+typedef void (*start_element)(void *, const XML_Char *, const XML_Char **);
+typedef void (*element_decl)(void *, const XML_Char *, XML_Content *);
+typedef void (*xml_decl)(
+    void *, const XML_Char *, const XML_Char *, int);
+typedef void (*start_doctype_decl)(
+    void *, const XML_Char *, const XML_Char *, const XML_Char *, int);
+typedef void (*notation_decl)(
+    void *,
+    const XML_Char *, const XML_Char *, const XML_Char *, const XML_Char *);
+typedef void (*attlist_decl)(
+    void *,
+    const XML_Char *, const XML_Char *, const XML_Char *, const XML_Char *,
+    int);
+typedef void (*unparsed_entity_decl)(
+    void *,
+    const XML_Char *, const XML_Char *,
+    const XML_Char *, const XML_Char *, const XML_Char *);
+typedef void (*entity_decl)(
+    void *,
+    const XML_Char *, int,
+    const XML_Char *, int,
+    const XML_Char *, const XML_Char *, const XML_Char *, const XML_Char *);
+typedef int (*external_entity_ref)(
+    XML_Parser,
+    const XML_Char *, const XML_Char *, const XML_Char *, const XML_Char *);
+
+#define SETTER_WRAPPER(NAME, TYPE)                                      \
+    static inline void                                                  \
+    pyexpat_Set ## NAME (XML_Parser parser, xmlhandler handler)         \
+    {                                                                   \
+        (void)XML_Set ## NAME (parser, (TYPE)handler);                  \
+    }
+
+SETTER_WRAPPER(StartElementHandler, start_element)
+SETTER_WRAPPER(EndElementHandler, parser_and_data)
+SETTER_WRAPPER(ProcessingInstructionHandler, parser_and_data_and_data)
+SETTER_WRAPPER(CharacterDataHandler, parser_and_data_and_int)
+SETTER_WRAPPER(UnparsedEntityDeclHandler, unparsed_entity_decl)
+SETTER_WRAPPER(NotationDeclHandler, notation_decl)
+SETTER_WRAPPER(StartNamespaceDeclHandler, parser_and_data_and_data)
+SETTER_WRAPPER(EndNamespaceDeclHandler, parser_and_data)
+SETTER_WRAPPER(CommentHandler, parser_and_data)
+SETTER_WRAPPER(StartCdataSectionHandler, parser_only)
+SETTER_WRAPPER(EndCdataSectionHandler, parser_only)
+SETTER_WRAPPER(DefaultHandler, parser_and_data_and_int)
+SETTER_WRAPPER(DefaultHandlerExpand, parser_and_data_and_int)
+SETTER_WRAPPER(NotStandaloneHandler, not_standalone)
+SETTER_WRAPPER(ExternalEntityRefHandler, external_entity_ref)
+SETTER_WRAPPER(StartDoctypeDeclHandler, start_doctype_decl)
+SETTER_WRAPPER(EndDoctypeDeclHandler, parser_only)
+SETTER_WRAPPER(EntityDeclHandler, entity_decl)
+SETTER_WRAPPER(XmlDeclHandler, xml_decl)
+SETTER_WRAPPER(ElementDeclHandler, element_decl)
+SETTER_WRAPPER(AttlistDeclHandler, attlist_decl)
+#if XML_COMBINED_VERSION >= 19504
+SETTER_WRAPPER(SkippedEntityHandler, parser_and_data_and_int)
+#endif
+#undef SETTER_WRAPPER
+
 static struct HandlerInfo handler_info[] = {
 
     // The cast to `xmlhandlersetter` is needed as the signature of XML
     // handler functions is not compatible with `xmlhandlersetter` since
     // their second parameter is narrower than a `const void *`.
 #define HANDLER_INFO(name) \
-    {#name, (xmlhandlersetter)XML_Set##name, my_##name},
+    {#name, (xmlhandlersetter)pyexpat_Set##name, (xmlhandler)my_##name},
 
     HANDLER_INFO(StartElementHandler)
     HANDLER_INFO(EndElementHandler)