fix(registry): ensure clean shutdown (#2901)

Avoid the error message:

```
INFO     2025-07-24 21:51:54,530 __main__:598 server: Received interrupt signal, shutting down gracefully...                                          
ERROR    2025-07-24 21:51:54,692 asyncio:1826 uncategorized: Task was destroyed but it is pending!                                                    
         task: <Task pending name='Task-15' coro=<refresh_registry() running at                                                                       
         /Users/leseb/Documents/AI/llama-stack/llama_stack/distribution/stack.py:356> wait_for=<Future pending cb=[Task.task_wakeup()]> cb=>  
```

llama_stack/distribution/server/server.py

@@ -56,6 +56,7 @@ from llama_stack.distribution.stack import (
     cast_image_name_to_string,
     construct_stack,
     replace_env_vars,
+    shutdown_stack,
     validate_env_pair,
 )
 from llama_stack.distribution.utils.config import redact_sensitive_fields
@@ -151,18 +152,7 @@ async def shutdown(app):
     Handled by the lifespan context manager. The shutdown process involves
     shutting down all implementations registered in the application.
     """
-    for impl in app.__llama_stack_impls__.values():
-        impl_name = impl.__class__.__name__
-        logger.info("Shutting down %s", impl_name)
-        try:
-            if hasattr(impl, "shutdown"):
-                await asyncio.wait_for(impl.shutdown(), timeout=5)
-            else:
-                logger.warning("No shutdown method for %s", impl_name)
-        except TimeoutError:
-            logger.exception("Shutdown timeout for %s ", impl_name, exc_info=True)
-        except (Exception, asyncio.CancelledError) as e:
-            logger.exception("Failed to shutdown %s: %s", impl_name, {e})
+    await shutdown_stack(app.__llama_stack_impls__)
 
 
 @asynccontextmanager