Bug 211745

Summary: Latest 5.11 git doesn't boot.
Product: Drivers Reporter: Robert M. Muncrief (rmuncrief)
Component: Video(DRI - non Intel)Assignee: drivers_video-dri
Status: RESOLVED INVALID    
Severity: blocking CC: alexdeucher
Priority: P1    
Hardware: x86-64   
OS: Linux   
Kernel Version: linux-git-5.11rc7.r141.g358feceebbf6 Subsystem:
Regression: No Bisected commit-id:
Attachments: Good dmesg
Good Xorg
Good journalctl
Bad dmesg
Bad Xorg
Bad journalctl

Description Robert M. Muncrief 2021-02-14 21:33:43 UTC
Created attachment 295271 [details]
Good dmesg

I started testing kernel 5.11 around rc5 and it has never booted. However the logs always seemed to be truncated, and I couldn't ssh into my system until today.  
  
However I compiled the latest git today andd ssh in and get useful information.  
  
Everything works great with 5.10.16, so I've attached dmesg_5.10.16_good.txt, Xorg.0.log_5.10.16_good.txt, and journalctl_5.10.16_good.txt to show what happens during a successful boot.  
  
And attached dmesg_5.11-git_fail.txt, Xorg.0.log_5.11-git_fail.txt, and journalctl_5.11-git_fail.txt to show what happens during the failed boot with 5.11 git.  
  
By the way, the journalctl log was produced with journalctl -b -k.
Comment 1 Robert M. Muncrief 2021-02-14 21:34:35 UTC
Created attachment 295273 [details]
Good Xorg
Comment 2 Robert M. Muncrief 2021-02-14 21:35:11 UTC
Created attachment 295275 [details]
Good journalctl
Comment 3 Robert M. Muncrief 2021-02-14 21:35:34 UTC
Created attachment 295277 [details]
Bad dmesg
Comment 4 Robert M. Muncrief 2021-02-14 21:36:15 UTC
Created attachment 295279 [details]
Bad Xorg
Comment 5 Robert M. Muncrief 2021-02-14 21:36:55 UTC
Created attachment 295281 [details]
Bad journalctl
Comment 6 Robert M. Muncrief 2021-02-14 21:38:50 UTC
I forgot to include the full git kernel name. It's linux-git-5.11rc7.r141.g358feceebbf6
Comment 7 Alex Deucher 2021-02-15 14:47:09 UTC
Can you bisect?
Comment 8 Robert M. Muncrief 2021-02-15 19:45:49 UTC
(In reply to Alex Deucher from comment #7)
> Can you bisect?  
  
I can bisect, but I updated and compiled linux-mainline with the final 5.11 from the Arch AUR this morning and my system boots just fine. I did a couple of quick tests playing videos, running VMs, etc. and everything looks good. So you can close this bug.  
  
I've done some system updates since yesterday so later today I'm going to compile the same linux-git again, and if it fails compile the current linux-git. If that fails then I'm going to compare  the linux-git and linux-mainline PKGBUILD files to see if the the error was caused by the linux-git build. I've used linux-git many times over the years and never had a problem with it but who knows. It's definitely worth checking out though.  
  
Thank you for your help and attention, and thanks to all the kernel devs for their awesome work. It is greatly appreciated.
Comment 9 Alex Deucher 2021-02-15 19:48:15 UTC
Only reporters can close kernel bugs.  Please close if this is not an issue.
Comment 10 Robert M. Muncrief 2021-02-15 19:55:40 UTC
(In reply to Alex Deucher from comment #9)
> Only reporters can close kernel bugs.  Please close if this is not an issue.

I've closed it as invalid since I'm not sure what the issue was and there was no specific code fix I'm aware of.