I don’t see anything immediately suspicious here, that last line with the gnome-shell error about meta_window_set_stack_position_no_sync is one I also see in my logs without having this problem.
Can you upload the full log here? You might have to do it as an upload, not an inline text, due to size.
One other thing I just remembered, look for entries of program crashes in the Problem Reporting (aka abrt) application. Maybe clear any existing problems there, force the issue again, and see if it reports anything new after getting back into the session.
This is a little surprising to me. What are the settings for:
Settings → Power → Screen Blank
Settings → Power → Automatic Suspend
Settings → Privacy & Security → (Blank Screen Delay, Automatic Screen Lock, Automatic Screen Lock Delay)
Settings → Power → Screen Blank
5 minutes
Settings → Power → Automatic Suspend
When on battery power
Delay: 15 Minutes
Plugged In: N/A
Settings → Privacy & Security → (Blank Screen Delay, Automatic Screen Lock, Automatic Screen Lock Delay)
Blank Screen Delay: 5 minutes
Automatic Screen Lock: enabled
Automatic Screen Lock Delay: Screen Turns Off
Lock Screen Notifications: enabled
As for uploading the full journalctl file from yesterday, it looks like the only authorized file formats for upload here are either image file formats or CAD file formats. I can just post it here unless there is another site you’d want me to upload it to?
The settings look essentially identical to mine - I use 15 instead of 5 minutes and no automatic suspend.
BTW I somehow had managed to miss that you’re on a FW16 whereas mine is a FW13 AMD.
Anyway: For the logs, maybe try an upload service like pastebin and post the URL here? There might be some quasi-private information in there (wireless LAN names, Bluetooth/WLAN MAC addresses, that type of thing) so you might want to edit those out before posting.
I am seeing this intermittently on a FW16. Will try flipping from power-options service back to ppd and see if I have a recurrence. I have not figured out how to trigger it, so will just have to wait for it to happen, or not happen.
Edit to add - arch, gnome, dGPU installed.
1 Like
Dimitris, here is a pastebin containing logs for the last time this error occurred earlier today.
Also small update, when my screen went black earlier I was NOT able to press CTRL + ALT + F3 to revive it like last time. I had to press and hold the power button to forcefully shut my laptop down to get things working again.
All this:
Oct 03 11:45:38 titan kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* dc_dmub_srv_log_diagnostic_data: DMCUB error - collecting diagnostic data
suggests something is indeed going on in the graphics stack. If you search for DMCUB there are some other threads that also reference this in graphics-related problems (not exactly the same error message based on my quick search).
I’d suggest a bug report in the upstream graphics driver repository/bugtracker. Also quick cc to @Mario_Limonciello from AMD here in case there’s something else to try/suggest.
Whaddayaknow, just happened to me. Upstream issue here.
As unfortunate as it is that you encountered this issue, I have to admit I’m glad it’s not just me. Have you had any more black screens since the first one? Searching for that error online has also returned results from others going back to early July of this year.
I want to say I’ve had similar “black screen on wakeup” events, but I haven’t kept any notes/into.
I also don’t usually walk away from an unlocked screen so I may be giving this too few chances to occur.
How reliably do you reproduce this?
So far, not reliably at all. There are times when it seems to do this immediately once the screen goes dark and locks due to inactivity, other times (like now) it won’t give me any issues for days on end.
There’s firmware updates after the 2024-09-09 date that corresponds to (currently) latest releases in various distros. New linux-firmware versions are cut roughly monthly, so we’ll hopefully get a chance to test this soon.
Here’s an interesting update, I have had multiple instances in the past few days where my entire session became incredibly slow and choppy while watching Youtube. During one of these slowdowns I was able to open System Monitor but nothing stood out. CPU, RAM, GPU, nothing was even close to being capped out. During the most recent slow down, I stepped away long enough for my system to lock due to inactivity and when I came back I had the black screen issue. I didn’t mention this at first because I didn’t know if it was related, but now I think they are. Had you been watching anything on your laptop prior to when you had your last black screen?
I can’t say yes or no for sure, but leaning towards no. I’ll try to keep a better note of when I watch videos (which isn’t that often).
I ran into another severe slow down, again while trying to watch something on Youtube. I did find in another thread (here) that running the command sudo cat /sys/kernel/debug/dri/1/amdgpu_gpu_recover will at least get things working again without rebooting, although it resets your session so be sure to save everything before running it. Just wanted to let you know in case you start to experience the slow down issue in addition to the black screen.
Quick note that the firmware updates are rolling out, here’s the Fedora update, currently still in updates-testing with:
amdgpu: DMCUB DCN35 update
2 Likes
Thanks for the heads up. I updated earlier today and noticed amdgpu firmware listed in the software to be updated so let’s hope this resolves the issue for everyone. I’ll post an update within a few days or sooner depending on how things are running.
I also just ran into a black screen, although I was lucky in that pressing CTRL + ALT + F3 and then switching back to F2 fixed it. Sometimes doing this works, sometimes it doesn’t. Oddly enough I do not see any amdgpu errors when I run journalctl this time around.
I just ran into this myself (Fedora 40, latest update, so kernel 6.11.3). I thought it was a fluke and thought nothing of it until I read this thread.
Running into the same issue with Ubuntu 24.04 and updated to the latest rc2 kernel update, but it didn’t solve the issue sadly