You are not logged in.

#1 2023-07-08 16:37:21

redhawk
Member
Registered: 2023-07-08
Posts: 5

Mircocode crashes on amd ryzen 3rd gen

The message
```
Jul 08 12:20:28 archlinux kernel: microcode: CPU1: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU0: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU2: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU4: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU3: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU5: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU6: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU7: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU9: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: mce: [Hardware Error]: Machine check events logged
Jul 08 12:20:28 archlinux kernel: microcode: CPU10: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU11: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: mce: [Hardware Error]: CPU 1: Machine Check: 0 Bank 5: bea0000000000108
Jul 08 12:20:28 archlinux kernel: mce: [Hardware Error]: TSC 0 ADDR 142ab25b6
Jul 08 12:20:28 archlinux kernel: fbcon: Taking over console
Jul 08 12:20:28 archlinux kernel: MISC d012000100000000 SYND 4d000000 IPID 500b000000000
Jul 08 12:20:28 archlinux kernel: mce: [Hardware Error]: PROCESSOR 2:870f10 TIME 1688833226 SOCKET 0 APIC 2 microcode 8701030
Jul 08 12:20:28 archlinux kernel: microcode: CPU8: patch_level=0x08701030
Jul 08 12:20:28 archlinux kernel: microcode: Microcode Update Driver: v2.2.

```

I've been trying to fix this issue many times, what I've tried.

* disabling cstates in my bios since people have said its linked to that

* using these kernel params `processor.max_cstate=5 rcu_nocbs=0-11`

Here's my full system specs https://mystb.in/AustriaProcurementHong
and just ask for anything else

edit
So after looking into the motherboard side of things I found this fedora user with the exact same problem as me.

https://discussion.fedoraproject.org/t/ … 600x/71833

Unlike him I updated my bios weeks ago so I shouldn't be having the issues https://www.asus.com/motherboards-compo … ME-B450M-A

Last edited by redhawk (2023-07-08 21:04:03)

Offline

#2 2023-07-08 17:07:52

frostschutz
Member
Registered: 2013-11-15
Posts: 1,646

Re: Mircocode crashes on amd ryzen 3rd gen

mce: and microcode: messages might be unrelated and just happen to appear at the same time

Offline

#3 2023-07-08 17:18:21

redhawk
Member
Registered: 2023-07-08
Posts: 5

Re: Mircocode crashes on amd ryzen 3rd gen

frostschutz wrote:

mce: and microcode: messages might be unrelated and just happen to appear at the same time

as what else though? I didn't mention it in the main post but I also have another computer with ryzen 1st gen i use as a server that has this problem, and each computer runs entirely different software.

Offline

#4 2023-07-08 17:31:04

loqs
Member
Registered: 2014-03-06
Posts: 18,884

Re: Mircocode crashes on amd ryzen 3rd gen

If you edit the boot options so the microcode is not updated (either by adding the option dis_ucode_ldr or removing the initrd entry for the microcode bundle) is the MCE still produced?  The MCE references bank 5,  is the module in that bank known to be good?

Last edited by loqs (2023-07-08 17:37:01)

Offline

#5 2023-07-08 17:42:42

redhawk
Member
Registered: 2023-07-08
Posts: 5

Re: Mircocode crashes on amd ryzen 3rd gen

loqs wrote:

If you edit the boot options so the microcode is not updated (either by adding the option dis_ucode_ldr or removing the initrd entry for the microcode bundle) is the MCE still produced?  The MCE references bank 5,  is the module in that bank known to be good?

im unsure how i would see if a bank is good or not. But this is a history of both mircocode errors i've got

└$ journalctl | grep "Hardware Error"
Jul 06 12:51:37 archlinux kernel: mce: [Hardware Error]: Machine check events logged
Jul 06 12:51:37 archlinux kernel: mce: [Hardware Error]: CPU 6: Machine Check: 0 Bank 5: bea0000000000108
Jul 06 12:51:37 archlinux kernel: mce: [Hardware Error]: TSC 0 ADDR 142ab2566 MISC d012000100000000 SYND 4d000000 IPID 500b000000000
Jul 06 12:51:37 archlinux kernel: mce: [Hardware Error]: PROCESSOR 2:870f10 TIME 1688662294 SOCKET 0 APIC 1 microcode 8701030
Jul 08 12:20:28 archlinux kernel: mce: [Hardware Error]: Machine check events logged
Jul 08 12:20:28 archlinux kernel: mce: [Hardware Error]: CPU 1: Machine Check: 0 Bank 5: bea0000000000108
Jul 08 12:20:28 archlinux kernel: mce: [Hardware Error]: TSC 0 ADDR 142ab25b6
Jul 08 12:20:28 archlinux kernel: mce: [Hardware Error]: PROCESSOR 2:870f10 TIME 1688833226 SOCKET 0 APIC 2 microcode 8701030

Offline

#6 2023-07-08 17:47:40

loqs
Member
Registered: 2014-03-06
Posts: 18,884

Re: Mircocode crashes on amd ryzen 3rd gen

redhawk wrote:
loqs wrote:

The MCE references bank 5,  is the module in that bank known to be good?

im unsure how i would see if a bank is good or not.

Have you run memtest on the system?
Edit:
If you use lshw to map the bank to a slot on the motherboard then swap the module in that slot with another module does the error still occur?  If so is it for the same bank or has the bank changed to follow the memory module?
Also is there a message "microcode updated early to new patch_level ..."  if not the kernel is only reporting the detected microcode reversion for each core and the timing would appear to be a coincidence.

Last edited by loqs (2023-07-08 18:43:08)

Offline

#7 2023-07-08 20:52:02

redhawk
Member
Registered: 2023-07-08
Posts: 5

Re: Mircocode crashes on amd ryzen 3rd gen

loqs wrote:
redhawk wrote:
loqs wrote:

The MCE references bank 5,  is the module in that bank known to be good?

im unsure how i would see if a bank is good or not.

Have you run memtest on the system?
Edit:
If you use lshw to map the bank to a slot on the motherboard then swap the module in that slot with another module does the error still occur?  If so is it for the same bank or has the bank changed to follow the memory module?
Also is there a message "microcode updated early to new patch_level ..."  if not the kernel is only reporting the detected microcode reversion for each core and the timing would appear to be a coincidence.


The error occurs rarely enough to where I couldn't test something like that without weeks going by. I did run memtest off the arch installer with 2 passes.

Offline

#8 2023-07-08 21:40:15

seth
Member
From: Won't reply 2 private help req
Registered: 2012-09-03
Posts: 75,652

Re: Mircocode crashes on amd ryzen 3rd gen

https://www.google.com/search?q=%22Bank … 0000108%22

Let's maybe  look at the actual problem, what are the symptoms when you experience a crash?
Freeze, reboot, …?

Did you see https://wiki.archlinux.org/title/Ryzen#Troubleshooting ?

Offline

#9 2023-07-08 22:25:13

redhawk
Member
Registered: 2023-07-08
Posts: 5

Re: Mircocode crashes on amd ryzen 3rd gen

seth wrote:

https://www.google.com/search?q=%22Bank … 0000108%22

Let's maybe  look at the actual problem, what are the symptoms when you experience a crash?
Freeze, reboot, …?

Did you see https://wiki.archlinux.org/title/Ryzen#Troubleshooting ?

What happens is a freeze on a black screen, I just set up zenstates https://aur.archlinux.org/packages/zenstates-git so hopefully that will help. However I am a little scared to mess with the voltages of anything that arent stock. I have never had good experiences changing those values.

Offline

Board footer

Powered by FluxBB