如何启用 Nvidia 编码器以及为什么卸载 Cuda 后出现 Nvidia 驱动程序错误

如何启用 Nvidia 编码器以及为什么卸载 Cuda 后出现 Nvidia 驱动程序错误

我的操作系统是 Ubuntu 18.04。

我想启用 NVENC 编码器进行 OBS 录制。我遵循了以下教程:https://scottlinux.com/2016/09/12/how-to-enable-nvidia-nvenc-for-obs-in-linux/

但是在第 2 步,我犯了一个错误。我从 ubuntu 存储库安装了 Cuda,而不是仅仅复制 Cuda.h。

我从存储库 (apt get cuda) 安装 Cuda 后,重新启动了笔记本电脑,但无法启动 GUI。卸载 Cuda 后,我可以再次启动 GUI,但卸载 cuda 时我也丢失了 nvidia 驱动程序。

然后,我尝试根据本指南重新安装 nvidia 驱动程序:https://www.cyberciti.biz/faq/ubuntu-linux-install-nvidia-driver-latest-proprietary-driver/ 我的 vga 是 Nvidia 930MX,并且我安装了 nvidia-driver-435。

安装驱动程序时没有任何错误。但是当我重新启动笔记本电脑时,我无法再次启动 GUI。我尝试了另一个版本 nvidia-driver-390、nvidia-driver-430。但没有任何效果。

现在,我决定使用 Intel HD 来应急。但在启动时(dmesg 日志),我收到此错误:

[  281.674620] pcieport 0000:00:1c.5: AER: Corrected error received: 0000:00:1c.5
[  281.674640] pcieport 0000:00:1c.5: AER: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)
[  281.674655] pcieport 0000:00:1c.5: AER:   device [8086:9d15] error status/mask=00001000/00002000
[  281.674667] pcieport 0000:00:1c.5: AER:    [12] Timeout               
[  307.377696] pcieport 0000:00:1c.5: AER: Corrected error received: 0000:00:1c.5
[  307.377708] pcieport 0000:00:1c.5: AER: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)
[  307.377718] pcieport 0000:00:1c.5: AER:   device [8086:9d15] error status/mask=00001000/00002000
[  307.377726] pcieport 0000:00:1c.5: AER:    [12] Timeout               
[  424.630009] systemd-journald[356]: File /var/log/journal/db71aa7f5c2f459083471d2a40ef5a91/user-1000.journal corrupted or uncleanly shut down, renaming and replacing.
[  425.928959] Bluetooth: RFCOMM TTY layer initialized
[  425.928968] Bluetooth: RFCOMM socket layer initialized
[  425.928978] Bluetooth: RFCOMM ver 1.11
[  427.172678] rfkill: input handler disabled
[  493.668193] pcieport 0000:00:1c.5: AER: Corrected error received: 0000:00:1c.5
[  493.668210] pcieport 0000:00:1c.5: AER: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)
[  493.668213] pcieport 0000:00:1c.5: AER:   device [8086:9d15] error status/mask=00003000/00002000
[  493.668214] pcieport 0000:00:1c.5: AER:    [12] Timeout               
[  501.530352] pcieport 0000:00:1c.5: AER: Corrected error received: 0000:00:1c.5
[  501.530358] pcieport 0000:00:1c.5: AER: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)
[  501.530360] pcieport 0000:00:1c.5: AER:   device [8086:9d15] error status/mask=00001000/00002000
[  501.530362] pcieport 0000:00:1c.5: AER:    [12] Timeout               
[  502.306367] userif-4: sent link down event.
[  502.306369] userif-4: sent link up event.
[  701.431701] pcieport 0000:00:1c.5: AER: Corrected error received: 0000:00:1c.5
[  701.431721] pcieport 0000:00:1c.5: AER: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)
[  701.431738] pcieport 0000:00:1c.5: AER:   device [8086:9d15] error status/mask=00001000/00002000
[  701.431749] pcieport 0000:00:1c.5: AER:    [12] Timeout               
[  771.912488] pcieport 0000:00:1c.5: AER: Corrected error received: 0000:00:1c.5
[  771.912509] pcieport 0000:00:1c.5: AER: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)
[  771.912522] pcieport 0000:00:1c.5: AER:   device [8086:9d15] error status/mask=00001000/00002000
[  771.912531] pcieport 0000:00:1c.5: AER:    [12] Timeout               

我想找回我的 nvidia 驱动程序。任何人都可以解决这个问题或给出建议的解决方案。

更新 1:现在我按照以下步骤解决了 nvidia 驱动程序错误问题。

  1. 删除所有 nvidia 包并重新启动
sudo apt purge nvidia-driver-440
sudo apt autoremove
  1. 从运行二进制文件安装 nvidia 驱动程序,可在nvidia 网站。. 如果出现 dkms 错误,请运行:
sudo apt install dkms

如果出现无法加载“nvidia-drm”内核模块的错误,请删除 xorg.conf

sudo rm /etc/X11/xorg.conf

然后安装二进制 nvida 驱动程序。

我正在等待您的帮助,以找到在 OBS 上启用 nvenc 编码器的最佳方法。

相关内容