Comment 562 for bug 1690085

(In reply to Borislav Petkov from comment #475)
> (In reply to Aaron Muir Hamilton from comment #474)
> > I just experienced this issue with a Threadripper 2950X on kernel 4.20.3
>
> Anything in dmesg?
>
> Please upload full dmesg, /proc/cpuinfo and kernel .config.
There is no header soldered on the board's UART, I do not have a normal serial port, and the logs would not write to disk (so I can't copy the dmesg from the incident). It amounted to:

BUG: soft lockup - CPU#13 stuck for 22s!

The config is here https://git.archlinux.org/svntogit/packages.git/tree/trunk/config?h=packages/linux

>
> Anything particular you did to reproduce it? Any correlation between what
> the box does and occurrence of the lockup?
Sadly no. The core happened to be running pulseaudio and upower, so my audio buffer started looping and eventually the watchdog timer timed out. I don't think any of that is particularly special though.

>
> Thx.

No, thank you! I don't know what I'm doing, and maybe you do. :- )

For the time being, I've now added idle=nomwait to my kernel params, as it seems at least one of these issues is related to MWAIT (according to AMD's errata).