Bug 106434

Summary: Graphical hangs, no keyboard, rest of system still active
Product: DRI Reporter: Koz Ross <koz.ross>
Component: DRM/AMDgpuAssignee: Default DRI bug account <dri-devel>
Status: RESOLVED MOVED QA Contact:
Severity: major    
Priority: medium CC: alpapad, harry.wentland
Version: unspecified   
Hardware: x86-64 (AMD64)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
Relevant fragment of dmesg when this occurs
none
Full dmesg
none
lshw
none
My CK patches for the 4.17-rc kernel none

Description Koz Ross 2018-05-08 04:30:44 UTC
Created attachment 139418 [details]
Relevant fragment of dmesg when this occurs

I sometimes (for seemingly no reason at all) have the graphics completely hang on my machine. I lose my keyboard as well, although my machine is still running (I can SSH into it and everything else still works). I can still move the mouse, but nothing visual responds (I can't click on anything, can't change active windows, etc). 

Some additional points of interest:

- My DisplayPort monitor blacks out in such a case, while my DVI monitor keeps working (displaying the same still image)
- Killing X or my WM has no effect on this once it occurs (only a restart fixes it)
- I can't determine if any particular thing triggers it (seems totally arbitrary)

I've attached a dmesg fragment from one such incident, as well as other pertinent information. Its mention of AMDGPU is why I reported this bug here.
Comment 1 Koz Ross 2018-05-08 04:31:38 UTC
Created attachment 139419 [details]
Full dmesg

Full dmesg of a normal start and a running system.
Comment 2 Koz Ross 2018-05-08 04:32:23 UTC
Created attachment 139420 [details]
lshw

lshw run as root
Comment 3 Koz Ross 2018-05-08 04:38:23 UTC
Created attachment 139421 [details] [review]
My CK patches for the 4.17-rc kernel

My mesa version is 18.1.0_rc3, libdrm-2.4.91, X.Org X Server 1.19.5. My kernel is 4.17-rc3, with the CK patches ported. I have attached the patch I applied to the kernel for completeness. I have disabled C-states in the BIOS, as well as ASLR and have rcu_nocb as a kernel parameter (due to Ryzen instability issues).
Comment 4 Alex Deucher 2018-05-09 03:37:16 UTC
Do you still have the issue without the CK patches?
Comment 5 Koz Ross 2018-05-09 06:14:34 UTC
Yes, I do - I tried using a stock kernel, as well as a CK one which wasn't 4.17-rc3, and they all had this behaviour.
Comment 6 Jack Wolf 2018-07-06 20:38:08 UTC
I have the same issue.

4.17.4 
firmware 20180518 
18.0.1 amdgpu

I also use displayport.
Comment 7 Martin Peres 2019-11-19 08:37:53 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/376.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.