gh-111784: Fix two segfaults in the elementtree module (GH-113405)

First fix resolve situation when pyexpat module (which contains expat_CAPI
capsule) deallocates before _elementtree, so we need to hold a strong
reference to pyexpat module to.

Second fix resolve situation when module state is deallocated before
deallocation of XMLParser instances, which uses module state to clear
some stuff.

1 files changed, 14 insertions, 2 deletions

diff --git a/Modules/_elementtree.c b/Modules/_elementtree.c
index f9d5793..5bf6787 100644
--- a/Modules/_elementtree.c
+++ b/Modules/_elementtree.c
@@ -98,6 +98,7 @@ typedef struct {
     PyTypeObject *TreeBuilder_Type;
     PyTypeObject *XMLParser_Type;
 
+    PyObject *expat_capsule;
     struct PyExpat_CAPI *expat_capi;
 } elementtreestate;
 
@@ -155,6 +156,7 @@ elementtree_clear(PyObject *m)
     Py_CLEAR(st->ElementIter_Type);
     Py_CLEAR(st->TreeBuilder_Type);
     Py_CLEAR(st->XMLParser_Type);
+    Py_CLEAR(st->expat_capsule);
 
     st->expat_capi = NULL;
     return 0;
@@ -175,6 +177,7 @@ elementtree_traverse(PyObject *m, visitproc visit, void *arg)
     Py_VISIT(st->ElementIter_Type);
     Py_VISIT(st->TreeBuilder_Type);
     Py_VISIT(st->XMLParser_Type);
+    Py_VISIT(st->expat_capsule);
     return 0;
 }
 
@@ -3066,6 +3069,7 @@ typedef struct {
     PyObject *handle_close;
 
     elementtreestate *state;
+    PyObject *elementtree_module;
 } XMLParserObject;
 
 /* helpers */
@@ -3607,7 +3611,11 @@ xmlparser_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
         self->handle_start = self->handle_data = self->handle_end = NULL;
         self->handle_comment = self->handle_pi = self->handle_close = NULL;
         self->handle_doctype = NULL;
-        self->state = get_elementtree_state_by_type(type);
+        self->elementtree_module = PyType_GetModuleByDef(type, &elementtreemodule);
+        assert(self->elementtree_module != NULL);
+        Py_INCREF(self->elementtree_module);
+        // See gh-111784 for explanation why is reference to module needed here.
+        self->state = get_elementtree_state(self->elementtree_module);
     }
     return (PyObject *)self;
 }
@@ -3784,6 +3792,7 @@ xmlparser_gc_clear(XMLParserObject *self)
         EXPAT(st, ParserFree)(parser);
     }
 
+    Py_CLEAR(self->elementtree_module);
     Py_CLEAR(self->handle_close);
     Py_CLEAR(self->handle_pi);
     Py_CLEAR(self->handle_comment);
@@ -4343,7 +4352,10 @@ module_exec(PyObject *m)
         goto error;
 
     /* link against pyexpat */
-    st->expat_capi = PyCapsule_Import(PyExpat_CAPSULE_NAME, 0);
+    if (!(st->expat_capsule = _PyImport_GetModuleAttrString("pyexpat", "expat_CAPI")))
+        goto error;
+    if (!(st->expat_capi = PyCapsule_GetPointer(st->expat_capsule, PyExpat_CAPSULE_NAME)))
+        goto error;
     if (st->expat_capi) {
         /* check that it's usable */
         if (strcmp(st->expat_capi->magic, PyExpat_CAPI_MAGIC) != 0 ||