[Feat] DiffusionEngine Support async batch infer (#2729)

Signed-off-by: Semmer <semmer@live.cn>
Signed-off-by: jader <yjader@foxmail.com>
Signed-off-by: asukaqaq-s <1311722138@qq.com>
Co-authored-by: asukaqaq-s <1311722138@qq.com>
Co-authored-by: jader <yjader@foxmail.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
Co-authored-by: Gao Han <hgaoaf@connect.ust.hk>

tests/diffusion/test_multiproc_engine_concurrency.py

@@ -19,6 +19,7 @@ from vllm_omni.diffusion.diffusion_engine import DiffusionEngine
 from vllm_omni.diffusion.executor.multiproc_executor import MultiprocDiffusionExecutor
 from vllm_omni.diffusion.sched import RequestScheduler
 from vllm_omni.diffusion.stage_diffusion_proc import StageDiffusionProc
+from vllm_omni.inputs.data import OmniDiffusionSamplingParams
 from vllm_omni.outputs import OmniRequestOutput
 
 pytestmark = [pytest.mark.diffusion, pytest.mark.core_model, pytest.mark.cpu]
@@ -34,7 +35,11 @@ def _tagged_output(tag: str) -> DiffusionOutput:
 
 def _mock_request(tag: str):
     """Return a lightweight request object identifiable by *tag*."""
-    return SimpleNamespace(request_ids=[tag])
+    return SimpleNamespace(
+        request_ids=[tag],
+        prompts=[f"prompt_{tag}"],
+        sampling_params=OmniDiffusionSamplingParams(num_inference_steps=1),
+    )
 
 
 def _make_executor(num_gpus: int = 1):
@@ -406,7 +411,7 @@ class TestCollectiveRpcTimeoutWhileLockHeld:
         t.start()
 
         # Wait until request execution is truly inside the lock and blocking.
-        add_req_blocked.wait(5)
+        assert add_req_blocked.wait(5)
 
         # collective_rpc should time out at lock acquisition, not hang.
         with pytest.raises(TimeoutError):