Stability Issues Ryzen 9 3900X, Aorus x570 Ultra MB, Linux 5.4

Hardware CPU
,
1

I recently bought and installed a Ryzen 9 3900X and Aorus X570 Ultra MB in my main linux desktop. It has been running mostly fine, but from time to time it will freeze, and then crash the enitre system.

Below are the errors that it spits out:

[kristoffer@TheOvermind ~]$ journalctl --since “yesterday” -p “emerg”…“crit”
– Logs begin at Mon 2020-01-27 22:57:27 CET, end at Sat 2020-02-22 00:13:18 CET. –
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: CPU 14: Machine Check: 0 Bank 5: bea0000000000108
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: TSC 0 ADDR 1f80202632010 MISC d012000100000000 SYND 4d000000 IPID 500b000000000
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: PROCESSOR 2:870f10 TIME 1582322485 SOCKET 0 APIC 5 microcode 8701013
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: CPU 20: Machine Check: 0 Bank 5: bea0000000000108
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: TSC 0 ADDR 6aefbf3a MISC d012000100000000 SYND 4d000000 IPID 500b000000000
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: PROCESSOR 2:870f10 TIME 1582322485 SOCKET 0 APIC 15 microcode 8701013
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: CPU 22: Machine Check: 0 Bank 5: bea0000000000108
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: TSC 0 ADDR 778aaf62 MISC d012000100000000 SYND 4d000000 IPID 500b000000000
Feb 21 23:01:28 TheOvermind kernel: mce: [Hardware Error]: PROCESSOR 2:870f10 TIME 1582322486 SOCKET 0 APIC 1b microcode 870101

Anyone have any clue what might be causing this?

2

Have you tried reseating the CPU and checked the pins?

How are the temps?

Also, what distro/kernel version?

There could be a number of things, but the most likely is going to be c-state if it’s not a hardware issue.

1 Like
3

Not tried reseating yet, was planning on trying that if it kept persisting.
Temps are fine, between 40 to 70 max.

Distro is Manjaro, kernel is 5.4
Previous CPU was Intel, so I made sure that I have amd-ucode installed so
i have up to date microcode loaded on startup.

4

Yeah, “disable c-states in BIOS” is my goto answer as well for any instability on Ryzen.

Other thing is memory, what RAM is in your box and what are the settings for it? Did you make sure it is in the board QVL?

2 Likes
5

Well, you’ve definitely done the needful as far as initial troubleshooting.

I’d remove any manual overclocks (on the CPU), disable c-states for now, and see if that helps.

If it’s still having issues, disable XMP and give that a go, once you have it stable, you can start turning things back on to see where your issue is.

2 Likes
6

Diabled c-states, I have had no problems so far, so is might have been just that. Thanks for the tips, let’s hope it stays stable.

2 Likes
7

Sorry to necro, just wanted to check in and see if your problems are still solved?

Recently did a fresh install of Ubuntu 20.04 from 18.04 and getting a ton of the same errors on my 3900X that was previously behaving fine under 18.04. No hardware changes.

8

For me it actually turned out in the end to be unstable memory clocks, which sucks since I was running my memory at 3200MHz which it was advertised as. Using a different XMP profile that actually underclocked it to 3000MHz solved all my instability issues.

Probably related to infinity fabric timings or something.

9

interesting, I am running my memory with the auto populated xmp profile at 2666mhz, so nothing crazy. Guess my cheapo B450 board just can’t handle it. Strange I had zero issues until I updated from Ubuntu 18.04 to 20.04 though. maybe a coincidence? Thanks for replying

10

This topic was automatically closed 273 days after the last reply. New replies are no longer allowed.