Bug 106317

Summary:

System lock up after monitors wake from DPMS on Ryzen 3 2200g

Product:

DRI

Reporter:

Kertesz Laszlo <laszlo.kertesz>

Component:

DRM/AMDgpu

Assignee:

Default DRI bug account <dri-devel>

Status:

RESOLVED MOVED

QA Contact:

Severity:

major

Priority:

medium

CC:

FD, harry.wentland

Version:

DRI git

Hardware:

x86-64 (AMD64)

OS:

Linux (All)

Whiteboard:

i915 platform:

i915 features:

Attachments:

Description	Flags
dmesg	none

Description Kertesz Laszlo 2018-04-30 04:51:49 UTC

I have a Ryzen 3 2200g CPU on a Gigabyte AB350M-HD3 mobo. I have 2 monitors connected (same model), one via HDMI, the other via DVI.

OS: Debian Testing, kernel compiled from mainline git (rc2+ now, after the rc3 DRM fixes patches merge). But this issue is present even with the kernel from 
https://cgit.freedesktop.org/~agd5f/linux/log/?h=drm-next-4.18-wip

The issue i encounter is the when the monitors enter dpms and tries to wake, the system freezes. 

This is not happening always, but it can be reproduced reliably with running 
"DISPLAY=:0 xset dpms force off"
Repeatedly. It will eventually lock up (minimum tries was 1, max 9).
One of the monitors (HDMI) will have the signal still cut, the other (DVI) will have a blank signal (this behavior seems consistent).

The system freezes totally (no ssh access, ping, magic keys etc). Also no usable logs. This is compounded by the issue of having to reboot repeatedly after this issue since the system will freeze at reboot when the KMS mode is enabled.

Maybe unrelated but i have this line in dmesg after each boot:

[    5.312547] [drm:construct [amdgpu]] *ERROR* construct: Invalid Connector ObjectID from Adapter Service for connector index:2! type 0 expected 3

xrandr output:

HDMI-A-0 connected primary 1440x900+1440+0 (normal left inverted right x axis y axis) 408mm x 255mm
   1440x900      59.89*+  74.98  
   1280x1024     75.02    60.02  
   1280x800      59.89  
   1152x864      75.00  
   1280x720      59.89  
   1024x768      75.03    70.07    60.00  
   832x624       74.55  
   800x600       72.19    75.00    60.32    56.25  
   640x480       75.00    72.81    66.67    59.94  
   720x400       70.08  
DVI-D-0 connected 1440x900+0+0 (normal left inverted right x axis y axis) 408mm x 255mm
   1440x900      59.89*+  74.98  
   1280x1024     75.02    60.02  
   1280x800      59.89  
   1152x864      75.00  
   1280x720      59.89  
   1024x768      75.03    70.07    60.00  
   832x624       74.55  
   800x600       72.19    75.00    60.32    56.25  
   640x480       75.00    72.81    66.67    59.94  
   720x400       70.08  
DisplayPort-0 disconnected (normal left inverted right x axis y axis)

Comment 1 Michel Dänzer 2018-04-30 09:00:04 UTC

Please attach the dmesg output, captured as late as possible.

Comment 2 Öyvind Saether 2018-04-30 14:14:04 UTC

This could be duplicate of https://bugs.freedesktop.org/show_bug.cgi?id=105018#c31

Comment 3 Kertesz Laszlo 2018-04-30 14:24:11 UTC

(In reply to Öyvind Saether from comment #2)
> This could be duplicate of
> https://bugs.freedesktop.org/show_bug.cgi?id=105018#c31

Not sure, when this happens to me the box is dead (no ssh etc).
Later i will try to dump "dmesg -wH" from a ssh session. I installed the dbg package and enabled the dc verbose logging option.

Comment 4 Kertesz Laszlo 2018-04-30 19:50:26 UTC

Created attachment 139234 [details]
dmesg

Comment 5 Kertesz Laszlo 2018-04-30 19:52:05 UTC

Tried it and there is absolutely no relevant output in dmesg.
After freeze i had to force-restart the computer 5 times to pass the amdgpu loading point.
Attached dmesg above.

Comment 6 Kertesz Laszlo 2018-06-01 18:52:43 UTC

I did some additional testing and it seems that if i disable the monitor connected through DVI (xrandr --output DVI-D-0 --off), the HDMI monitor has no issues with dpms.

Comment 7 Kertesz Laszlo 2018-06-01 19:02:51 UTC

Tested with the DVI monitor only (disabled the HDMI) too and seems to work fine aswell.

I also tested with both and even now with kernel 4.17.0-rc7+ it locks up.

Comment 8 Martin Peres 2019-11-19 08:37:20 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/370.

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.