机器基础配置：物理服务器为海光7390，安装8张Z100。虚拟云主机为64核心，128G内存，kylin-desktop-v10sp1操作系统。

已部署DTK环境：24.04.3

已部署anaconda环境，python3.8,

已安装DAS：torch-2.3.0+das.opt2.dtk24043-cp38-cp38-manylinux_2_28_x86_64.whl

问题：当挂载1张DCU时，rocm-smi, rocminfo, hy-smi,均正常输出。pytorch调用cuda正常。

当挂载8张DCU时，rocm-smi正常显示8张DCU，rocminfo报错，且usermod -a -G 39 gea将用户加入39组后仍无效。pytorch无法调用cuda。

报错文字：hsa api call failure at: /data/jenkins_workspace/workspace/galaxy_release/galaxy/rocminfo/rocminfo.cc:1165

Call returned HSA_STATUS_ERROR_OUT_OF_RESOURCES: The runtime failed to allocate the necessary resources. This error may also occur when the core runtime library needs to spawn threads or create internal OS-specific events.