Elik, Ed,
I have overlooked tracemalloc module raises deadlock if apply the patch.
I found out a source comment on Modules/_tracemalloc.c:161
/* If your OS does not provide native thread local storage, you can implement
it manually using a lock. Functions of thread.c cannot be used because
they use PyMem_RawMalloc() which leads to a reentrant call. */
#if !(defined(_POSIX_THREADS) || defined(NT_THREADS))
# error "need native thread local storage (TLS)"
#endif
Py_HAVE_NATIVE_TLS is used only on thread source code. Therefore C Compiler couldn't report error about tracemalloc. I'm sorry that I didn't check test.
Currently I'm trying to implement new API based on msg281227.