[3.15] gh-159739: Avoid destroying a remote unwinder cache twice (GH-159982) (#160023)

In a free-threaded build, if allocating the thread-local bytecode cache fails after the code-object cache has been created, the initializer destroys the first cache and returns an error without clearing its pointer. Failed construction then invokes the unwinder's deallocator, which destroys the same cache again and can crash the profiling process instead of propagating `MemoryError`.
(cherry picked from commit b577a54)

Co-authored-by: Jelle Zijlstra <jelle@openai.com>

Author: miss-islington

Misc/NEWS.d/next/Library/2026-10-20-00-00-00.gh-issue-159739.5d4269.rst

@@ -0,0 +1 @@
+Fix a double free if remote unwinder initialization runs out of memory.

Modules/_remote_debugging/module.c

@@ -450,7 +450,6 @@ _remote_debugging_RemoteUnwinder___init___impl(RemoteUnwinderObject *self,
         NULL
     );
     if (self->tlbc_cache == NULL) {
-        _Py_hashtable_destroy(self->code_object_cache);
         PyErr_NoMemory();
         set_exception_cause(self, PyExc_MemoryError, "Failed to create TLBC cache");
         return -1;