Graphics card might be freezing system

[ciphin78]

Hi All, First time posting. I'll try to be as detailed as I can....

I just built a new desktop PC (first new PC I've had in 10 years). It was a kit from newegg. Here are the components:

• EVGA 500 BA 100-BA-0500-K1 500W ATX12V / EPS12V SLI CrossFire 80 PLUS BRONZE Certified Non-Modular Active PFC Power Supply
  
  
• ASUS TUF B450-PLUS GAMING AM4 AMD B450 SATA 6Gb/s ATX AMD Motherboard
  
  
• ASRock Phantom Gaming D Radeon RX 570 DirectX 12 RX570 4G 4GB 256-Bit GDDR5 PCI Express 3.0 x16 HDCP Ready Video Card
  
  
• AMD RYZEN 5 2600X 6-Core 3.6 GHz (4.2 GHz Max Boost) Socket AM4 95W YD260XBCAFBOX Desktop Processor
  
  

Currently running Ubuntu 18.04.5 LTS, but I've also tried 20.04 & 20.10.
What is happening:

After a fresh OS install things appear to be up and running great for about 1 day. After that amount of time, the system will freeze. I will reboot and things work good for about an hour and then freeze. This can repeat forever.

I removed the RX570 card and put in an old MSI graphics card just to see if that helped and to my surprise the system is not suffering the freezing (this is what i'm running at the moment to post this).

I attempted to install the amdgpu drivers from the amd site for the 570 (https://www.amd.com/en/support/graphics/radeon-500-series/radeon-rx-500-series/radeon-rx-570) for 18.04.5. I thought that worked at first but then after a day the freezing started again.

I'm somewhat new to Linux. I've been looking at the logs but I don't really know what I'm looking for and nothing is jumping out at me....so, i'm hear asking for help.

I'm happy to perform any kind of diagnostics and/or log reporting, I just don't know what else to do at the moment - any direction would be great.

Let me know if more information about my system is needed, thanks!

[ciphin78]

Well, It would seem that the GPU may only be part of the problem. The system froze up again today. I'm running a Memtest and there are failures...not sure what to do about that at the moment.

[MartyBuntu]

If you have more than one stick of RAM installed, try testing them one at a time.

[MartyBuntu]

Also, your graphics card is bumping up against the very limit of what your power supply can serve. That may or may not be adding to your concerns.

[ciphin78]

Please expand about the graphics card and PS. The recommend PSU for the graphics card is 500Watt.

I removed the first stick of ram (I have 4) and reran memtest and it came back clean. So i put the rx570 back in and left out the first ram stick and i'm running the system now. We'll see if it locks up again.

Are there more hardware tests I should do?

Thanks!

[kurt18947]

I had something similar happen when I had an Asrock AB350M motherboard and VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Caicos PRO [Radeon HD 7450] GPU running 18.04. Installing HWE helped some but it still wasn't stable. I then installed 20.04 which was an improvement but still had occasional hard lockups. I tried a different AMD video card but no change in stability. Nothing out of the ordinary showed up in dmesg and I'm not really savvy with reading logs. I swapped out the motherboard to an MSI B450 based board and reinstalled Ubuntu 20.04.1. So far this seems to behave as it should. The video card is using the Radeon driver.

[Autodave]

If that runs clean right now, then I would strongly suspect that your bad RAM chip is the one not in the machine.

[mastablasta]

ryzen - 95 W
Radeon RX 570 150W
--
240 W
+ motherboard + disks+ram should bring you up to 300W-350 W on max

500W bronze should provide at least around 400 W so you should be on the safe side regarding power supply. just a quick calculation.

dmesg shows the errors from boot onwards. there should be xorg or wayland logs you can search for and logviewer will provide you with other logs and errors. you can see if something is wrong if it says ERROR next to the message.

psensor can be used to check for any overheating. lockups can also happen because of that. if PCu is overheating , then paste wasnt' applied correctly, or the cooler isn't sitting on CPU well.

RAM stick can also cause issues. memtest will show you that.

or it may just be that motherbroard is crappy.

[ciphin78]

Thanks. It seems like things are better, however Chrome seems to be locking up and crashing. I've turned on crash reporting so I can see what it says the reason is. There is a *.crash file in /var/crash for Chrome, but i don't have a good way to read it... If you know of a good way, please tell me. So, i think things are getting better but not tweaked to stable just yet.

[ciphin78]

Can you provide me with some more information about xorg / wayland logs?

I've switched back to the old video card for now. I'm using psensor and the CPU temp seems to be hovering between 26C - 40C which i think is okay? I've run memtest several times without the one ram stick and it keeps showing clear of errors.

Here's some info from LOGS:
-Applications: Cannot access vdagent virtio channel /dev/virtio-ports/com.redhat.spice.0
-Applications: [pulseaudio] bluez5-util.c: GetManageObjects() failed: org.freedesktop.DBus.Error.AcessDenied: Rejected sent message
-Applications: CRITICAL: We failed, but the fail whale is dead. Sorry....
-Applciations: [pulseaudio] client-conf-x11.c: xcb_connection_has_error() returned true
-Other [1307]: error: Failed to create /var/spool/cups/tmp/.hplip
-Other Unrecoverable failure in required component org.gnome.shell.desktop