make pp group initilized to avoid pooint-to-point communication as th…

…e first call

Signed-off-by: yisheng <[email protected]>

vllm/worker/xpu_worker.py

@@ -11,6 +11,7 @@
 from vllm.config import VllmConfig
 from vllm.distributed import (ensure_model_parallel_initialized,
                               init_distributed_environment)
+from vllm.distributed.parallel_state import get_pp_group
 from vllm.logger import init_logger
 from vllm.model_executor import set_random_seed
 from vllm.platforms import current_platform
@@ -176,3 +177,8 @@ def init_worker_distributed_environment(self) -> None:
             parallel_config.pipeline_parallel_size)
         # global all_reduce needed for overall oneccl warm up
         torch.distributed.all_reduce(torch.zeros(1).xpu())
+
+        if parallel_config.pipeline_parallel_size > 1:
+            # Add pp group init to avoid
+            # p2p communication as the first call
+            get_pp_group().all_reduce(torch.zeros(1).xpu())