Note I experimenting with liquorix kernel but that ultimately was not the cause of my boot hangs
Original Issue:
After installing the liquorix kernel to improve gaming performance, I noticed system stability issues like frequently failing to boot or crashing when playing a game like Cyberpunk.
Here's how I installed liquorix:
sudo add-apt-repository ppa:damentz/liquorix && sudo apt-get update
sudo apt-get install linux-image-liquorix-amd64 linux-headers-liquorix-amd64
It was fine at first but I started running into thes stability issues after:
- Updating Mesa GL using Ubuntu Software Updater
- Removing unused kernel files from linux-image-5.4.0-64-generic (as recommended by the ubuntu software updater) <-- This seems to have triggered the stability issues
I even tried booting into my previous kernel (5.4.0-70-generic) however the problems still persisted. I decided to remove the liquorix kernel using these instructions:
Still the problems persisted, I even tried:
- sudo apt-get autoremove
- Removing all old kernels with Ubuntu Cleaner
- Reinstalling my current kernel
Unfortunately Ubuntu hangs on boot/reboot about 50% of the time. I am not sure why my system is so unstable after reverting back to the previous kernel and purging the old kernel. Not sure exactly what caused this, was it:
- Installing the liquorix kernel
- Updating Mesa GL after installing the liquorix kernel
- Removing the old kernel files
Also here is my Mesa version:
OpenGL version string: 4.6 (Compatibility Profile) Mesa 21.0.2 - kisak-mesa PPA
When it fails to boot, sometimes it hangs on my Motherboard logo. Othertimes I see a list of errors like this:
0.316293] do_IRQ: 1.55 No irq handler for vector
0.316293] do_IRQ: 2.55 No irq handler for vector
0.316293] do_IRQ: 3.55 No irq handler for vector
0.316293] do_IRQ: 4.55 No irq handler for vector
0.316293] do_IRQ: 5.55 No irq handler for vector
0.316293] do_IRQ: 6.55 No irq handler for vector
0.316293] do_IRQ: 7.55 No irq handler for vector
0.316293] do_IRQ: 8.55 No irq handler for vector
0.316293] do_IRQ: 9.55 No irq handler for vector
0.316293] do_IRQ: 10.55 No irq handler for vector
Update 1 It seems like the No irq handler is a red herring. It sometimes boots with that message.
When I boot without "quiet splash" I see this when it freezes
fb: switching to amdgpudrmfb from EFI VGA
Note that when it succeeds to boot it also displays these same messages for a split second as well. I am not sure why it sometimes succeeds and sometimes fails.
Update 2 I managed to find a log of what was installed that caused this in the first place:
2021-04-09 15:21:50 install linux-headers-liquorix-amd64:amd64 <none> 5.11-26ubuntu1~focal
2021-04-09 15:21:50 install linux-image-liquorix-amd64:amd64 <none> 5.11-26ubuntu1~focal
2021-04-09 15:27:15 upgrade libdrm-common:all 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:15 upgrade libdrm2:i386 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm2:amd64 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm-amdgpu1:amd64 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm-amdgpu1:i386 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm-intel1:i386 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm-intel1:amd64 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm-nouveau2:amd64 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm-nouveau2:i386 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm-radeon1:amd64 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libdrm-radeon1:i386 2.4.102-1ubuntu1kisak1~f 2.4.103-1ubuntu1~f~mesarc0
2021-04-09 15:27:16 upgrade libegl-mesa0:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:16 upgrade libgbm1:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:17 upgrade libgl1-mesa-dri:i386 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:17 upgrade libgl1-mesa-dri:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:18 upgrade libosmesa6:i386 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:18 upgrade libosmesa6:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libglx-mesa0:i386 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libglx-mesa0:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libglapi-mesa:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libglapi-mesa:i386 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libegl1-mesa:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libegl1-mesa-dev:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libgl1-mesa-dev:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libgl1-mesa-glx:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade libxatracker2:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:19 upgrade linux-firmware:all 1.187.10 1.190.2~mesarc0
2021-04-09 15:27:25 upgrade mesa-va-drivers:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:26 upgrade mesa-vdpau-drivers:i386 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:26 upgrade mesa-vdpau-drivers:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:26 upgrade mesa-vulkan-drivers:i386 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:27 upgrade mesa-vulkan-drivers:amd64 21.0.1~kisak1~f 21.0.2~kisak1~f
2021-04-09 15:27:27 upgrade xserver-xorg-video-amdgpu:amd64 19.1.0-1 19.1.0+git2102021245.6ed4863~f~mesarc0
2021-04-09 15:59:01 remove linux-headers-5.4.0-66-generic:amd64 5.4.0-66.74 <none>
2021-04-09 15:59:01 remove linux-headers-5.4.0-66:all 5.4.0-66.74 <none>
2021-04-09 15:59:02 remove linux-modules-extra-5.4.0-66-generic:amd64 5.4.0-66.74 <none>
2021-04-09 15:59:03 remove linux-image-5.4.0-66-generic:amd64 5.4.0-66.74 <none>
2021-04-09 15:59:07 remove linux-modules-5.4.0-66-generic:amd64 5.4.0-66.74 <none>
I've looked up various sources and it seems like the amdgpu firmware might be the culprit. I've reinstalled my current kernel and also reinstalled the firmware, even going so far as to delete the /lib/firmware/amdgpu and forcing a reinstallation. The problem is still there and I am not sure what else to do at this point.
Any ideas?