Did you try re-seating (and possibly cleaning) the RX560?
Was the PSU sufficiently powered (at least equal to) the previous PSU in wattage?
Yes, it was thoroughly cleaned on Saturday, I re-seated it multiple times, even switching the motherboard, to no avail. The new PSU is 900W, it should be well above the previous one.
More new information.
OK. It’s best if you list everything you’ve done; and in a detailed fashion; don’t gloss over anything. This will help greatly, if anyone is to help.
Also, please specify which monitors you are using, their resolutions, and which ports they are connected to.
The game of 1000 questions is rarely productive.
Please post the output from:
journalctl -p err -b
This might help in the meantime.
First I would try to see what process uses GPU so intensely that it heats up to 77 C in desktop mode. nvtop might help you with that.
Second may be try to boot other Linux distributives from live iso (without actually installing them) to see if GPU heats up to the same degree.
Aaaaaaa I’m sorry ;-;
It’s just a lot has happened over a short time, and I haven’t really been asking for help for a long time :P.
So, in the beginning, I’ve had my little original rig of Asus B450 mobo, with RX560 and Ryzen 5 1600, with some small 450W (I believe?) PSU, on which I’ve been playing games under Manjaro for some years now. Everything worked fairly fine, until like a month ago, I’ve been having an issue with my rig powering down, or at least the screens switching down on better days. It was fine, it seemed more stable on a single screen, but not entirely solving my issue. I’ve been considering the PSU to fail, so when I’ve had a chance to nab the Dell Precision 5820 tower on thursday (my current spec) I went for that. This hasn’t helped anything.
On Saturday I thoroughly cleaned and reapplied thermal paste to the GPU, thinking it would solve the problem. It didn’t. I started looking around some other solutions, and discovered on Sunday that everything seems working OK when testing under my Windows partition.
The command hasn’t returned anything particularly useful:
paź 23 21:08:40 superfortress systemd-modules-load[414]: Failed to find module 'acpi_call'
paź 23 21:08:42 superfortress kernel: EDAC skx: ECC is disabled on imc 0
paź 23 21:08:45 superfortress kernel:
paź 23 21:08:48 superfortress systemd[980]: Failed to start Update XDG user dir configuration.
paź 23 21:10:39 superfortress pulseaudio[1463]: GetManagedObjects() failed: org.freedesktop.DBus.Error.TimedOut: Failed to activate service 'org.bluez': timed out (service_start_timeout=25000ms)
Nothing in particular does, just a bit of memory usage from kwin_x11 and Firefox. I will try checking out with some live ISO, maybe it will lead me to some conclusion if it’s Manjaro’s drivers or something else.
If I understand correctly, it crashes with two monitors, and does not crash with only one monitor connected. True?
-
Which ports are they connected to?
-
What type of monitors are they?
-
Are they high resolution (eg: 2560x1440)
-
Do you have another Graphic card you could try instead?
Possible you not aware, but thats alone if heavily unprecise information.
Did you measured your temps before and after?
Are you aware, that you need sometimes more than less Termal Paste specially at some bad GPU Coolers? So its possible you can make it even worse… Im not saying that this is actually the case, but your weak information, left everything possible.
Well, I have places to be.
Good luck 
DELL P2419H, 1920x1080 displayport
BENQ GW2280 1920x1090 HDMI
Yes and no. I’ve got the Quadro P1000, but it has 4 miniDP outputs to which I don’t have cables. I can run native Linux stuff with prime-run decently without everything bombing out. Wine stuff doesn’t launch, but that’s a different can of worms specific to itself.
Just quickly, before I ride off into the sunset…
Have you tried using the DVI output instead of the DP?
I haven’t saved my findings because I haven’t thought about it, but it seemed fairly consistent with the ones I have right now.
I wasn’t aware that I might need more paste, but thanks to my friend I got a high quality (14W/m2) product, so even if I screwed it up, I don’t think this is an issue.
And I don’t think I did, seeing how I get stable temperatures under Windows and no poweroffs.
Nope, don’t have the cable for that. Oh, and also I didn’t answer the question: it does still happen under single (HDMI or DP) monitor plugged. It just isn’t instant, and happens in 2-15minutes of heavy load.
OK. I’ll come back to this next time, if nobody else finds a solution.
Cheers.
Thanks for trying, I really appreciate it!
1 Like
I think I have it:
The issue why on Windows it has been working perfectly and didn’t overheat is that the game itself was running under the second GPU, as somehow Windows utilized it as High Performance one that should be turned on during gameplay. As soon as I manually ordered it to run the game under RX560 - poof, system crash.
It seems like the poor RX560 is just faulty, and it’s not an issue with kernel, drivers or anything by itself. It simply outlived it’s usefulness.
Press F for the trusty card in my name, for we have conquered many foes together.
Thanks @Kobold, @Xephon and @soundofthunder for your time and brainpower.
Signing off!
1 Like
So … you have a dual-gpu setup and have not been using the dGPU?
What driver profile is installed?
mhwd -l -li
Assuming its PRIME then you want(ed) to be using prime-run, ex:
glxinfo | grep 'renderer string'
prime-run glxinfo | grep 'renderer string'
> Installed PCI configs:
--------------------------------------------------------------------------------
NAME VERSION FREEDRIVER TYPE
--------------------------------------------------------------------------------
video-linux 2018.05.04 true PCI
video-hybrid-amd-nvidia-prime 2023.03.23 false PCI
Warning: No installed USB configs!
> 0000:91:00.0 (0300:1002:67ef) Display controller ATI Technologies Inc:
--------------------------------------------------------------------------------
NAME VERSION FREEDRIVER TYPE
--------------------------------------------------------------------------------
video-hybrid-amd-nvidia-prime 2023.03.23 false PCI
video-hybrid-amd-nvidia-470xx-prime 2023.03.23 false PCI
video-linux 2018.05.04 true PCI
video-modesetting 2020.01.13 true PCI
video-vesa 2017.03.12 true PCI
> 0000:04:00.0 (0300:10de:1cb1) Display controller nVidia Corporation:
--------------------------------------------------------------------------------
NAME VERSION FREEDRIVER TYPE
--------------------------------------------------------------------------------
video-hybrid-amd-nvidia-prime 2023.03.23 false PCI
video-hybrid-amd-nvidia-470xx-prime 2023.03.23 false PCI
video-nvidia 2023.03.23 false PCI
video-nvidia-470xx 2023.03.23 false PCI
video-nvidia-390xx 2023.03.23 false PCI
video-linux 2018.05.04 true PCI
video-modesetting 2020.01.13 true PCI
video-vesa 2017.03.12 true PCI
OpenGL renderer string: AMD Radeon RX 560 Series (polaris11, LLVM 16.0.6, DRM 3.54, 6.6.0-1-MANJARO)
OpenGL renderer string: Quadro P1000/PCIe/SSE2
The caveat is that prime-run is working okay for native apps, but somehow clashes with Wine. I’ve already raised an issue on their forums, but at this point I think the easiest option would be just to buy a correct pair of cables and use the Quadro as my primary rig.
This could be a number of things…
https://wiki.archlinux.org/title/PRIME#Black_screen_with_GL-based_compositors
https://wiki.archlinux.org/title/PRIME#Error_when_running_Wine_games_with_DXVK
(note that if you do want to use one of the switcher things … best to use one of these: More tools for Optimus laptops are now available)
1 Like
This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.