[Perf] Zero-copy torch.Tensor pickling in shm_broadcast MessageQueue (#48442)

Signed-off-by: Ruinan Ma <r7ma3088@gmail.com> Signed-off-by: Nick Hill <nickhill123@gmail.com> Co-authored-by: Nick Hill <nickhill123@gmail.com>
2026-07-28 16:28:28 -07:00
parent bb3b61f2fd
commit a07fac758f
3 changed files with 233 additions and 3 deletions
@@ -48,6 +48,7 @@ CHECK_IMPORTS = {
            "vllm/distributed/device_communicators/shm_object_storage.py",
            "vllm/distributed/weight_transfer/ipc_engine.py",
            "vllm/distributed/weight_transfer/clients.py",
+            "tests/distributed/test_shm_broadcast.py",
            "tests/distributed/test_weight_transfer.py",
            "vllm/utils/hashing.py",
            "tests/multimodal/media/test_base.py",