NVIDIA卡---A100的驱动在22.03-LTS环境安装时报错

环境信息
1.驱动地址：https://developer.download.nvidia.com/compute/cuda/11.6.0/local_installers/cuda_11.6.0_510.39.01_linux_sbsa.run
2. 解压：sh cuda_11.6.0_510.39.01_linux_sbsa.run --noexec --target cuda
3. cd cuda/builds
4. 运行驱动：./NVIDIA-Linux-aarch64-510.47.03.run
5. 运行后报错文件： nvidia-install.log

环境信息

Hi hu-xuejiao, welcome to the openEuler Community.
I'm the Bot here serving you. You can find the instructions on how to interact with me at Here.
If you have any questions, please contact the SIG: Kernel, and any of the maintainers: @YangYingliang , @pi3orama , @成坚 (CHENG Jian) , @Qiuuuuu , @zhengzengkai , @gogooo , @Xie XiuQi

补充信息：
架构：aarch64
操作系统版本：openEuler 22.03LTS
内核版本：5.10.0

报错信息：
FATAL: modpost: GPL-incompatible module nvidia-drm.ko uses GPL-only symbol 'pv_ops'

NVIDIA-Linux-aarch64-510.47.03驱动非GPL兼容,但要求使用GPL导出的pv_ops变量
pv_ops在arch/arm64/kernel/paravirt.c定义并导出

struct paravirt_patch_template pv_ops = {
#ifdef CONFIG_PARAVIRT_SPINLOCKS
         .lock.queued_spin_lock_slowpath = native_queued_spin_lock_slowpath,
         .lock.queued_spin_unlock                = native_queued_spin_unlock,
#endif
         .lock.vcpu_is_preempted         = __native_vcpu_is_preempted,
};
EXPORT_SYMBOL_GPL(pv_ops);

pv_ops为aarch64架构为华为自研特性，可以同步社区x86架构的修改方案：

x86架构pv_ops修改为EXPORT_SYMBOL(pv_ops)
https://lore.kernel.org/all/20181029150116.25372-1-jgross@suse.com/T/#u

合入补丁解决：
353e2d0407 arm/arm64: paravirt: Remove GPL from pv_ops export

GVP openEuler / kernel

内容风险标识

评论 (5)

GVPopenEuler / kernel

内容风险标识