Comment 273 for bug 1690085

Revision history for this message
In , bluescarni (bluescarni-linux-kernel-bugs) wrote :

(In reply to eric.c.morgan from comment #187)
> Run the python script mentioned earlier.

+1, disabling C6 has resulted in a rock-solid experience over here, no idle reboots in months of constant use of the machine.

Just wanted to mention a couple of points for those still following this thread.

I recently updated my motherboard's BIOS to the latest version, which contains a CPU microcode update from AMD (I am now on version "0x08001136", whereas I was stuck on version "0x08001129" for - I believe - 6 months or so). I had high hopes a microcode update would fix the idle reboots, but, alas, they still persist with the new microcode and C6 enabled.

The other interesting tidbit is this article from 1 year ago I found:

https://www.gamersnexus.net/news-pc/2870-ryzen-power-plan-update-min-frequency-90-pct

Apparently, soon after the ryzen launch last year, AMD issued an updated version of the Windows "Balanced" preset power plan which disables "core parking". Doing a bit of research, it seems like core parking is synonymous (or, at least strongly related) to the C6 power state. Might this explain why people on Windows are not experiencing random reboots? I am not sure 100% how these Windows preset power plans work, but the page linked above seems to indicate that:

1) after the update, the "Balanced" power plan disables core parking,
2) core parking was always disabled anyway in the "Performance" power plan.

See also this thread on the AMD community forum:

https://community.amd.com/thread/220563

Maybe AMD knew all along there are issues with the C6 power state...