PCIE Bus Error: Severity=Corrected after booting into Ubuntu 16.04.2 LTS

Graphics / Linux Linux
#1

Setup a new server with 2x Nvidia K40 GPU Cards and 2x Xeon E5-2640 V4 and successfully install Ubuntu 16.04.2 LTS.

Next i keep getting the above error.

Though i have update the server system bios to the latest bios as well as install Ubuntu 16.04.3 LTS but the problem still occurs

The above error is getting on my nerve as this was the first time i’m getting the above error after setting up numerous servers running on ubuntu LTS with nvidia GPU cards

Pls assist

#2

Do kernel parameters pci=nomsi or pci=noaer help?
Other than that, run nvidia-bug-report.sh and attach output file.

#3

Thanks for dropping by

I did a search and did came across similar thread as following linked

Once i did a sudo gedit /etc/default/grub, it say command couldn’t be found

End up i abandon the Do kernel parameters pci=nomsi

by

GRUB_CMDLINE_LINUX_DEFAULT="quiet splash pci=nomsi"

Save and close gedit window.

Update grub and restart your system:

sudo update-grub
sudo reboot
#4

Care to share how do we going about adding to the grub pci=nomsi and pci=noaer?

Pls bear with me as i’m a noob when comes to ubuntu command

#5

The howto is correct, it just uses gedit which is the Gnome Texteditor. Which doesn’t exist if you’re not running gnome. Instead, use nano:
sudo nano /etc/default/grub
make the changes
CTRL+o then ENTER to save, CTRL+x to exit.
then run
sudo update-grub
and reboot.

#6

Are you able to provide me the exact step and command from Do kernel parameters pci=nomsi till sudo nano /etc/default/grub make the changes CTRL+o then ENTER to save, CTRL+x to exit.

And run sudo update-grub and reboot.

#7

I’m able to log into Nano and make the changes as follow:

GRUB_CMDLINE_LINUX_DEFAULT=“quiet splash pci=nomsi”

and save the file with: Ctrl+o, then Enter

Close nano with: Ctrl+x

next update grub: sudo update-grub

Follow by reboot your system

But i was welcome with a blank screen instead of allowing me to log in instead :-(

#8

Use CTRL+F1 to get to a text console, login and revert the changes, if you don’t already have.
If you can’t, hold shift key on startup to get to the grub menu, then press e to edit the comandline, use cursors to move to the change, delete it, then press F10 to boot.

If you can login again, then run
sudo nvidia-bug-report.sh
and attach the file it creates to your post.

#9

After i make the changes to GRUB_CMDLINE_LINUX_DEFAULT=“quiet splash pci=nomsi”

And next reboot I’m only able to see the following error and image for less than a few secs and next the whole screen turn blank

displaying ubuntu 16.04 with power meter ACPI000D:00: Ignoring unsafe software power cap image

#10

I was also facing the same issue, the mentioned trick doesn’t work in some devices like mine. In such a case you may struggle a lot like i did. But eventually i found a video through which i fixed this issue https://youtu.be/kBMRw7j4tlg