GitLab

Hi, I have an ASUS ROG Zephyrus G14 with AMD Ryzen 7 4800HS, the iGPU shows as this:

04:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Renoir (rev c6)

When I try to run superposition on a external monitor connected to HDMI output:

Screen 0: minimum 320 x 200, current 3840 x 1200, maximum 16384 x 16384
eDP connected primary 1920x1080+0+0 (normal left inverted right x axis y axis) 309mm x 174mm
   1920x1080    120.00*+  60.00  
   1680x1050    120.00  
   1280x1024    120.00  
   1440x900     120.00  
   1280x800     120.00  
   1280x720     120.00  
   1024x768     120.00  
   800x600      120.00  
   640x480      120.00  
HDMI-A-0 connected 1920x1200+1920+0 (normal left inverted right x axis y axis) 518mm x 324mm
   1920x1200     59.95*+
   1920x1080     60.00  
   1600x1200     60.00  
   1680x1050     59.88  
   1280x1024     60.02  
   1440x900      59.90  
   1280x960      60.00  
   1280x800      59.95  
   1280x720      59.95  
   1024x768      60.00  
   800x600       60.32    56.25  
   640x480       59.94  

It immediately crashes, the system is still accessible via ssh, crash log:

[  104.394534] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring sdma0 timeout, signaled seq=8257, emitted seq=8259
[  104.394584] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process launcher pid 1497 thread launcher:cs0 pid 1565
[  104.394590] amdgpu 0000:04:00.0: GPU reset begin!
[  104.412252] [drm] [Link 0] WARNING MOD_HDCP_STATUS_DISPLAY_NOT_FOUND IN STATE HDCP_UNINITIALIZED STAY COUNT 0
[  104.428687] [drm] [Link 0] WARNING MOD_HDCP_STATUS_DISPLAY_NOT_FOUND IN STATE HDCP_UNINITIALIZED STAY COUNT 0
[  104.555360] amdgpu 0000:04:00.0: GPU reset succeeded, trying to resume
[  104.555555] [drm] PCIE GART of 1024M enabled (table at 0x000000F400900000).
[  104.555785] [drm] PSP is resuming...
[  104.555790] sos fw version = 0x110049.
[  104.575820] [drm] reserve 0x400000 from 0xf41f800000 for PSP TMR
[  104.884742] SMU is resuming...
[  104.884808] SMU is resumed successfully!
[  105.083853] [drm] kiq ring mec 2 pipe 1 q 0
[  105.232962] amdgpu 0000:04:00.0: [drm:amdgpu_ring_test_helper [amdgpu]] *ERROR* ring sdma0 test failed (-110)
[  105.233004] [drm:amdgpu_device_ip_resume_phase2 [amdgpu]] *ERROR* resume of IP block <sdma_v4_0> failed -110
[  105.233047] amdgpu 0000:04:00.0: GPU reset(1) failed
[  105.233061] amdgpu 0000:04:00.0: GPU reset end with ret = -110
[  115.274545] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring sdma0 timeout, signaled seq=8259, emitted seq=8259
[  115.274602] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process launcher pid 1497 thread launcher:cs0 pid 1565
[  115.274608] amdgpu 0000:04:00.0: GPU reset begin!

What is super strange, superposition works perfectly fine when starting it from the build in display on the laptop. And this even when the HDMI monitor is connected, just need to start it on the eDP. Kernel is 5.7.2 now. From time to time I also experience those crashes doing normal desktop things (but only when the HDMI monitor is connected), but they are not often (maybe once ever 8h).

BTW I also tried amdgpu.noretry=0 and it still happens