操作系统：ubuntu

内核版本：6.8.04

DTK版本：26.04  

在启动VLLM时报错，内如如下：

WARN[0000] /home/docker/docker-compose.yml: the attribute `version` is obsolete, it will be ignored, please remove it to avoid potential confusion

Attaching to vllm-qwen3.8-offline

vllm-qwen3.8-offline-1  | get hyhal driver version error!

vllm-qwen3.8-offline-1  | INFO 09-01 22:33:40 [__init__.py:44] Available plugins for group vllm.platform_plugins:

vllm-qwen3.8-offline-1  | INFO 09-01 22:33:40 [__init__.py:46] - hcu -> vllm_hcu:hcu_platform_plugin

vllm-qwen3.8-offline-1  | INFO 09-01 22:33:40 [__init__.py:49] All plugins in this group will be loaded. Set `VLLM_PLUGINS` to control which plugins to load.

vllm-qwen3.8-offline-1  | INFO 09-01 22:33:40 [__init__.py:239] Platform plugin hcu is activated

vllm-qwen3.8-offline-1  | Traceback (most recent call last):

vllm-qwen3.8-offline-1  |   File "/usr/local/bin/vllm", line 3, in <module>

vllm-qwen3.8-offline-1  |     from vllm.entrypoints.cli.main import main

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm/entrypoints/cli/__init__.py", line 3, in <module>

vllm-qwen3.8-offline-1  |     from vllm.entrypoints.cli.benchmark.latency import BenchmarkLatencySubcommand

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm/entrypoints/cli/benchmark/latency.py", line 5, in <module>

vllm-qwen3.8-offline-1  |     from vllm.benchmarks.latency import add_cli_args, main

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm/benchmarks/latency.py", line 16, in <module>

vllm-qwen3.8-offline-1  |     from vllm.engine.arg_utils import EngineArgs

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/arg_utils.py", line 35, in <module>

vllm-qwen3.8-offline-1  |     from vllm.config import (

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm/config/__init__.py", line 6, in <module>

vllm-qwen3.8-offline-1  |     from vllm.config.compilation import (

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm/config/compilation.py", line 22, in <module>

vllm-qwen3.8-offline-1  |     from vllm.platforms import current_platform

vllm-qwen3.8-offline-1  |   File "<frozen importlib._bootstrap>", line 1075, in _handle_fromlist

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm/platforms/__init__.py", line 279, in __getattr__

vllm-qwen3.8-offline-1  |     _current_platform = resolve_obj_by_qualname(platform_cls_qualname)()

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm/utils/import_utils.py", line 111, in resolve_obj_by_qualname

vllm-qwen3.8-offline-1  |     module = importlib.import_module(module_name)

vllm-qwen3.8-offline-1  |   File "/usr/lib/python3.10/importlib/__init__.py", line 126, in import_module

vllm-qwen3.8-offline-1  |     return _bootstrap._gcd_import(name[level:], package, level)

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm_hcu/platforms/__init__.py", line 1, in <module>

vllm-qwen3.8-offline-1  |     from .hcu import HCUPlatform

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm_hcu/patch_utils.py", line 149, in _custom_import

vllm-qwen3.8-offline-1  |     module = OLD_IMPORT_HOOK(

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm_hcu/platforms/hcu.py", line 78, in <module>

vllm-qwen3.8-offline-1  |     _ON_GFX938 = "gfx938" in _get_gcn_arch_name()

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/vllm_hcu/platforms/hcu.py", line 74, in _get_gcn_arch_name

vllm-qwen3.8-offline-1  |     GPU_ARCH = torch.cuda.get_device_properties("cuda").gcnArchName

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/torch/cuda/__init__.py", line 632, in get_device_properties

vllm-qwen3.8-offline-1  |     _lazy_init()  # will define _get_device_properties

vllm-qwen3.8-offline-1  |   File "/usr/local/lib/python3.10/dist-packages/torch/cuda/__init__.py", line 424, in _lazy_init

vllm-qwen3.8-offline-1  |     torch._C._cuda_init()

vllm-qwen3.8-offline-1  | RuntimeError: No HIP GPUs are available

vllm-qwen3.8-offline-1 exited with code 1