AMD Zen 3 (5800X) support

[oxgl]

Hi,

during system boot I'm getting these error messages:

EDAC amd64: F19h detected (node 0).
EDAC amd64: Error: F0 not found, device 0x1650 (broken BIOS?)

I'm not experiencing any issues with stability.
My configuration:
CPU: AMD Ryzen 7 5800X
MB: ASUS ROG STRIX X570-E GAMING
GPU: MSI GT 1030 2GH LP OC
RAM: G.SKILL 32GB KIT DDR4 3600MHz CL16 Trident Z RGB

Should I ignore these messages? What does they mean? Is it because the Zen 3 is not (yet) supported by the kernel?

Thank you!

oxgl

[Miuku]

Leap's kernel is too old for it to properly support 5k series unless they backported things, I would use the Kernel:/Stable (5.10.7 at this moment) if you have Ryzen 5000.

[oxgl]

Thank you! Does it mean I can change the kernel in Leap? How can I do that? Will it work without any issues - this is my production PC.

[Miuku]

Thank you! Does it mean I can change the kernel in Leap? How can I do that? Will it work without any issues - this is my production PC.

If you have no issues other than the warnings - just ignore them and wait for 15.3 :-)

[oxgl]

If you have no issues other than the warnings - just ignore them and wait for 15.3 :-)

Yes, but this page writes that Leap 15.3 will also use kernel 5.3.18, so it means 2 years from now I will see those warnings? I've tried to upgrade the kernel version based on some manual, but after rebooting the same kernel was running, and I was not able to start some applications (Chrome, Firefox). Could it be a problem that now it's running on "5.3.18-lp152.57-preempt" kernel?

[Miuku]

Yes, but this page writes that Leap 15.3 will also use kernel 5.3.18, so it means 2 years from now I will see those warnings? I've tried to upgrade the kernel version based on some manual, but after rebooting the same kernel was running, and I was not able to start some applications (Chrome, Firefox). Could it be a problem that now it's running on "5.3.18-lp152.57-preempt" kernel?

Preempt is aimed at very specific workloads and not for your average use - it can cause issues with applications and normal desktop users should not use it.

However, if you are feeling adventurous you could add the Kernel devel repo:

https://download.opensuse.org/repositories/Kernel:/stable/standard/

and install kernel-default from there.

Note; you should keep your previous stable 15.2 kernel installed before you adventure into latest-hot-stuff land :-) Also this will cause issues with nVidia proprietary drivers so you should not use this repo if you use the pre-built nVidia drivers.

[oxgl]

Thank you! I've uninstalled preempt kernel. I don't know when or why it was installed - maybe related to Virtualbox installation? I don't know. Anyway I've changed it to default kernel, then I've installed version 5.10, and the message is gone. I'm using the proprietary nVidia driver 460.32.03 with this kernel, but no issues yet. But I'm not (yet) playing games on this machine (it's just a GeForce GT 1030 ) What kind of issues should I experience?

There are still these irq messages:

2021-01-15T00:08:03.849511+01:00 install kernel: [    0.496441] .... node  #0, CPUs:        #1
2021-01-15T00:08:03.849511+01:00 install kernel: [    0.316206] __common_interrupt: 1.55 No irq handler for vector
2021-01-15T00:08:03.849511+01:00 install kernel: [    0.496482]   #2
2021-01-15T00:08:03.849511+01:00 install auditd[1007]: Started dispatcher: /sbin/audispd pid: 1010
2021-01-15T00:08:03.849513+01:00 install kernel: [    0.316206] __common_interrupt: 2.55 No irq handler for vector
2021-01-15T00:08:03.849513+01:00 install kernel: [    0.498577]   #3
2021-01-15T00:08:03.849513+01:00 install kernel: [    0.316206] __common_interrupt: 3.55 No irq handler for vector
2021-01-15T00:08:03.849513+01:00 install auditd[1007]: Init complete, auditd 2.8.1 listening for events (startup state enable)
2021-01-15T00:08:03.849513+01:00 install kernel: [    0.500482]   #4
2021-01-15T00:08:03.849514+01:00 install kernel: [    0.316206] __common_interrupt: 4.55 No irq handler for vector
2021-01-15T00:08:03.849515+01:00 install kernel: [    0.502577]   #5
2021-01-15T00:08:03.849515+01:00 install audispd: No plugins found, exiting
2021-01-15T00:08:03.849516+01:00 install kernel: [    0.316206] __common_interrupt: 5.55 No irq handler for vector
2021-01-15T00:08:03.849516+01:00 install kernel: [    0.504482]   #6
2021-01-15T00:08:03.849516+01:00 install kernel: [    0.316206] __common_interrupt: 6.55 No irq handler for vector
2021-01-15T00:08:03.849517+01:00 install kernel: [    0.506577]   #7
2021-01-15T00:08:03.849517+01:00 install kernel: [    0.316206] __common_interrupt: 7.55 No irq handler for vector
2021-01-15T00:08:03.849516+01:00 install auditd[1007]: dispatcher 1010 reaped
2021-01-15T00:08:03.849517+01:00 install kernel: [    0.508483]   #8
2021-01-15T00:08:03.849518+01:00 install kernel: [    0.316206] __common_interrupt: 8.55 No irq handler for vector
2021-01-15T00:08:03.849518+01:00 install apparmor.systemd[694]: Warning from stdin (line 1): Cache: failed to add read only location '/usr/share/apparmor/cache', does not contain valid cache directory for the specified feature set
2021-01-15T00:08:03.849519+01:00 install kernel: [    0.510613] Spectre V2 : Update user space SMT mitigation: STIBP always-on
2021-01-15T00:08:03.849519+01:00 install kernel: [    0.510613]   #9
2021-01-15T00:08:03.849519+01:00 install kernel: [    0.316206] __common_interrupt: 9.55 No irq handler for vector
2021-01-15T00:08:03.849520+01:00 install kernel: [    0.512482]  #10
2021-01-15T00:08:03.849520+01:00 install kernel: [    0.316206] __common_interrupt: 10.55 No irq handler for vector
2021-01-15T00:08:03.849520+01:00 install kernel: [    0.514574]  #11 #12 #13 #14 #15

Is this a problem? Same on kernel 5.3 and also on 5.10.

[arvidjaar]

2021-01-15T00:08:03.849511+01:00 install kernel: [    0.496441] .... node  #0, CPUs:        #1
2021-01-15T00:08:03.849511+01:00 install kernel: [    0.316206] __common_interrupt: 1.55 No irq handler for vector

Is this a problem?

It is pretty popular and nobody reported any issues associated with it. So far common agreement is that it is probably BIOS doing something stupid. There was brief discussion on LKML.

https://lkml.org/lkml/2019/2/19/516

[OrsoBruno]

...then I've installed version 5.10, and the message is gone. I'm using the proprietary nVidia driver 460.32.03 with this kernel, but no issues yet. But I'm not (yet) playing games on this machine (it's just a GeForce GT 1030 ) What kind of issues should I experience?

The Nvidia kmps available in the Leap 15.2 repo are tailored for 5.3.x kernels and will _not_work with 5.10.x kernels, but if you installed the Nvidia driver "the hard way" and you rebuilt the modules for 5.10.x you should be good to go.

[oxgl]

It is pretty popular and nobody reported any issues associated with it. So far common agreement is that it is probably BIOS doing something stupid. There was brief discussion on LKML.

https://lkml.org/lkml/2019/2/19/516

So I should simply ignore it! Thank you!