[PD] Introduce runtime role switching between prefill and decode (#28403)

Signed-off-by: huanglong <huanglong@linux.alibaba.com>
Signed-off-by: inkcherry <mingzhi.liu@amd.com>
Co-authored-by: huanglong <huanglong@linux.alibaba.com>
Co-authored-by: Shangming Cai <csmthu@gmail.com>
Co-authored-by: Huang Long <121648372+LLLL114@users.noreply.github.com>

test/registered/unit/entrypoints/test_server_info.py

@@ -69,6 +69,7 @@ class TestModelInfoSerialization(CustomTestCase):
             "load_format": _CustomModelLoader,
             "reasoning_parser": None,
             "tool_call_parser": None,
+            "disaggregation_mode": "null",
         }
         tokenizer_manager = SimpleNamespace(
             model_config=SimpleNamespace(
@@ -97,6 +98,7 @@ class TestModelInfoSerialization(CustomTestCase):
             reset_context()
 
         self.assertEqual(payload["load_format"], f"{__name__}._CustomModelLoader")
+        self.assertEqual(payload["disaggregation_mode"], "null")
         json.dumps(payload)