A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 6, 17:00 UTC
[Bugfix] Avoid LD_PRELOAD check on MacOS (#37145)
Signed-off-by: jiang1.li <jiang1.li@intel.com>
Li, Jiang committed
Mar 16, 2026 at 14:31 UTC
7362b4450a4cde8b208682c0be6c901d4b5290e6
1 file changed
+5
-3
vllm/v1/worker/cpu_worker.py
+5
-3
@@ -2,6 +2,7 @@
2
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
3
import os
4
import platform
5
+import sys
6
from collections.abc import Callable
7
from typing import Any
8
@@ -63,9 +64,10 @@ class CPUWorker(Worker):
64
"to setup required pre-loaded libraries."
65
)
66
66
- check_preloaded_libs("libtcmalloc")
67
- if current_platform.get_cpu_architecture() == CpuArchEnum.X86:
68
- check_preloaded_libs("libiomp")
67
+ if sys.platform.startswith("linux"):
68
+ check_preloaded_libs("libtcmalloc")
69
+ if current_platform.get_cpu_architecture() == CpuArchEnum.X86:
70
+ check_preloaded_libs("libiomp")
71
72
# Setup OpenMP threads affinity.
73
omp_cpuids = envs.VLLM_CPU_OMP_THREADS_BIND