AMDGPU error: "[drm:amdgpu_dm_atomic_commit_tail [amdgpu]] *ERROR* Waiting for fences timed out!"

[Michel Dänzer]

On 2021-07-11 9:48 a.m., Ketsui wrote:
>> So far, so good; no hang in a week. I'll try the rest of the new firmware as well now, will follow up if there's a hang again.
> 
> I've noticed that the VM_L2_PROTECTION_FAULT_STATUS error doesn't always result in a hang, looking through my journal I can see
> maybe a dozen of them spread out across multiple boots but my system only became non-functional like two times so far (I know
> because I have the dmesg when the hangs happened saved, you can find them attached to this mail).
> 
> To make myself clear, I haven't actually had a hang too with the new firmwares even though these messages still appear on my dmesg,
> sorry if my feedback gave the wrong impression.

I'm counting soft recovered hangs as hangs for the purpose of this issue. I.e. when I write "no hang" I mean no soft recovered ones either. If I hit a soft recovered hang, I consider that setup bad.


-- 
Earthling Michel Dänzer               |               https://redhat.com
Libre software enthusiast             |             Mesa and X developer