pythongh-116522: Stop the world before fork() and during shutdown

This changes the free-threaded build to perform a stop-the-world pause
before deleting other thread states when forking and during shutdown.
This fixes some crashes when using multiprocessing and during shutdown
when running with `PYTHON_GIL=0`.

This also changes `PyOS_BeforeFork` to acquire the runtime lock
(i.e., `HEAD_LOCK(&_PyRuntime)`) before forking to ensure that data
protected by the runtime lock (and not just the GIL or stop-the-world)
is in a consistent state before forking.

Modules/posixmodule.c

@@ -613,11 +613,16 @@ PyOS_BeforeFork(void)
     run_at_forkers(interp->before_forkers, 1);
 
     _PyImport_AcquireLock(interp);
+    _PyEval_StopTheWorldAll(&_PyRuntime);
+    HEAD_LOCK(&_PyRuntime);
 }
 
 void
 PyOS_AfterFork_Parent(void)
 {
+    HEAD_UNLOCK(&_PyRuntime);
+    _PyEval_StartTheWorldAll(&_PyRuntime);
+
     PyInterpreterState *interp = _PyInterpreterState_GET();
     if (_PyImport_ReleaseLock(interp) <= 0) {
         Py_FatalError("failed releasing import lock after fork");
@@ -632,6 +637,7 @@ PyOS_AfterFork_Child(void)
     PyStatus status;
     _PyRuntimeState *runtime = &_PyRuntime;
 
+    // re-creates runtime->interpreters.mutex (HEAD_UNLOCK)
     status = _PyRuntimeState_ReInitThreads(runtime);
     if (_PyStatus_EXCEPTION(status)) {
         goto fatal_error;
@@ -7858,9 +7864,9 @@ os_fork1_impl(PyObject *module)
         /* child: this clobbers and resets the import lock. */
         PyOS_AfterFork_Child();
     } else {
-        warn_about_fork_with_threads("fork1");
         /* parent: release the import lock. */
         PyOS_AfterFork_Parent();
+        warn_about_fork_with_threads("fork1");
     }
     if (pid == -1) {
         errno = saved_errno;
@@ -7906,9 +7912,9 @@ os_fork_impl(PyObject *module)
         /* child: this clobbers and resets the import lock. */
         PyOS_AfterFork_Child();
     } else {
-        warn_about_fork_with_threads("fork");
         /* parent: release the import lock. */
         PyOS_AfterFork_Parent();
+        warn_about_fork_with_threads("fork");
     }
     if (pid == -1) {
         errno = saved_errno;
@@ -8737,9 +8743,9 @@ os_forkpty_impl(PyObject *module)
         /* child: this clobbers and resets the import lock. */
         PyOS_AfterFork_Child();
     } else {
-        warn_about_fork_with_threads("forkpty");
         /* parent: release the import lock. */
         PyOS_AfterFork_Parent();
+        warn_about_fork_with_threads("forkpty");
     }
     if (pid == -1) {
         return posix_error();

Python/pylifecycle.c

@@ -1918,6 +1918,9 @@ Py_FinalizeEx(void)
     runtime->initialized = 0;
     runtime->core_initialized = 0;
 
+    /* Ensure that remaining threads are detached */
+    _PyEval_StopTheWorldAll(runtime);
+
     // XXX Call something like _PyImport_Disable() here?
 
     /* Destroy the state of all threads of the interpreter, except of the

Python/pystate.c

@@ -1715,6 +1715,10 @@ _PyThreadState_DeleteExcept(PyThreadState *tstate)
     PyInterpreterState *interp = tstate->interp;
     _PyRuntimeState *runtime = interp->runtime;
 
+#ifdef Py_GIL_DISABLED
+    assert(runtime->stoptheworld.world_stopped);
+#endif
+
     HEAD_LOCK(runtime);
     /* Remove all thread states, except tstate, from the linked list of
        thread states.  This will allow calling PyThreadState_Clear()
@@ -1733,6 +1737,8 @@ _PyThreadState_DeleteExcept(PyThreadState *tstate)
     interp->threads.head = tstate;
     HEAD_UNLOCK(runtime);
 
+    _PyEval_StartTheWorldAll(runtime);
+
     /* Clear and deallocate all stale thread states.  Even if this
        executes Python code, we should be safe since it executes
        in the current thread, not one of the stale threads. */