From c48e03f6aee0df8467588f98f1b7ed418a2ec409 Mon Sep 17 00:00:00 2001 From: trtllm-agent <296075020+trtllm-agent@users.noreply.github.com> Date: Thu, 16 Jul 2026 03:37:31 -0700 Subject: [PATCH] [nvbugs/6463819][fix] Use is_comm_session() to check pool session in proxy The concrete-type guard `isinstance(self.mpi_session, MpiPoolSession)` in GenerationExecutorProxy._start_executor_workers raises TypeError when the test-suite session-reuse infrastructure rebinds tensorrt_llm.executor.proxy.MpiPoolSession to a factory function and hands out _ReusableSession wrappers (not subclasses). Replace with the polymorphic predicate `not self.mpi_session.is_comm_session()` already defined on MpiSession; it delegates through the wrapper's __getattr__, and preserves the original semantics (pool sessions register worker PIDs; comm sessions skip). Signed-off-by: trtllm-agent <296075020+trtllm-agent@users.noreply.github.com> --- tensorrt_llm/executor/proxy.py | 2 +- tests/integration/test_lists/waives.txt | 1 - 2 files changed, 1 insertion(+), 2 deletions(-) diff --git a/tensorrt_llm/executor/proxy.py b/tensorrt_llm/executor/proxy.py index 11451bad5b67..8aa7a6e524ed 100644 --- a/tensorrt_llm/executor/proxy.py +++ b/tensorrt_llm/executor/proxy.py @@ -573,7 +573,7 @@ def mpi_done_callback(future: concurrent.futures.Future): raise RuntimeError( "Executor worker returned error") from ready_signal - if isinstance(self.mpi_session, MpiPoolSession) and len(status) == 3: + if not self.mpi_session.is_comm_session() and len(status) == 3: worker_process_identities: List[WorkerProcessIdentity] = status[2] self._worker_process_monitor.register(worker_process_identities) diff --git a/tests/integration/test_lists/waives.txt b/tests/integration/test_lists/waives.txt index 4b6a59b069ec..1c4de79288df 100644 --- a/tests/integration/test_lists/waives.txt +++ b/tests/integration/test_lists/waives.txt @@ -457,7 +457,6 @@ unittest/_torch/ray_orchestrator/multi_gpu/test_llm_update_weights_multi_gpu.py unittest/_torch/ray_orchestrator/multi_gpu/test_llm_update_weights_multi_gpu.py::test_llm_partial_update_weights_nvfp4[auto-Qwen3/Qwen3-8B] SKIP (https://nvbugs/6372690) unittest/_torch/ray_orchestrator/multi_gpu/test_llm_update_weights_multi_gpu.py::test_llm_partial_update_weights_nvfp4[fp8-Qwen3/Qwen3-30B-A3B] SKIP (https://nvbugs/6372690) unittest/_torch/ray_orchestrator/multi_gpu/test_llm_update_weights_multi_gpu.py::test_llm_partial_update_weights_nvfp4[fp8-Qwen3/Qwen3-8B] SKIP (https://nvbugs/6372690) -unittest/_torch/sampler/test_beam_search.py::test_beam_search_e2e[multi_process-TRTLLMSampler-cuda_graph_and_overlap-None-1-1-True-True-False] SKIP (https://nvbugs/6463819) unittest/_torch/speculative/hw_agnostic/test_dflash.py::test_dflash_qwen3_5_4b[False] SKIP (https://nvbugs/6451032) unittest/_torch/speculative/hw_agnostic/test_dflash.py::test_dflash_qwen3_5_4b[True] SKIP (https://nvbugs/6451032) unittest/_torch/speculative/test_eagle3.py::test_llama_eagle3[True-TRTLLM-True-False-True-True-True-False-False-False] SKIP (https://nvbugs/6451425)