search

jianyunchao / mindlearning

MindSpore learning from scratch.
0 stars 0 forks source link

【nvidia】【driver】ubuntu系统内核自动更新导致驱动读取失败 #10

Open jianyunchao opened 11 months ago

jianyunchao commented 11 months ago

背景：执行nvidia-smi显示报错 NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running. 解决思路：查看内核ls /boot，发现内核升级导致原有驱动不可用，回退到历史的内核版本 or 重新安装driver驱动和固件适配

jianyunchao commented 11 months ago

image

jianyunchao commented 11 months ago

查看系统是否有多个内核

ls -l /boot image

jianyunchao commented 11 months ago

查找启动配置文件

find /boot/ -name "grub.cfg"

image

jianyunchao commented 11 months ago

image

jianyunchao commented 11 months ago

将历史196内核替换，复制196两行替换为第二个export下面的，注意不要选到到recovery后缀的，为紧急模式

image

jianyunchao commented 11 months ago

重启系统

reboot

uname -r查看系统版本，再执行nvidia-smi已经显示正常 image

jianyunchao commented 11 months ago

禁止内核自动升级

将apt自动升级设置为0，关闭状态 vim /etc/apt/apt.conf.d/10periodic vim /etc/apt/apt.conf.d/20auto-upgrades image image

apt-mark hold $(uname -r) 查看： dpkg --get-selections | grep hold