-
Notifications
You must be signed in to change notification settings - Fork 6
Thermal/power management settings not restored on S3 resume path #1795
Description
Component
Dasharo firmware
Device
NovaCustom V54 14th Gen
Dasharo version
v1.0.0
Dasharo Tools Suite version
No response
Test case ID
No response
Brief summary
Debugging of QubesOS/qubes-issues#10616 lead to finding that after resume Xen complains about CPU thermals:
(XEN) CPU0: Temperature above threshold
(XEN) CPU0: Running in modulated clock mode
(XEN) CPU0: Temperature/speed normal
(XEN) CPU0: Temperature above threshold
(XEN) CPU0: Running in modulated clock mode
(XEN) CPU0: Temperature/speed normal
(XEN) CPU0: Temperature above threshold
(XEN) CPU0: Running in modulated clock mode
(XEN) CPU0: Temperature/speed normal
(XEN) CPU0: Temperature above threshold
(XEN) CPU0: Running in modulated clock mode
(XEN) CPU0: Temperature/speed normal
(XEN) CPU0: Temperature above threshold
(XEN) CPU0: Running in modulated clock mode
And it remains in that state - note there is no final CPU0: Temperature/speed normal, it remains in "modulated clock mode", at least according to Xen. Physically, the laptop is cold, it's fresh after resuming and idle. No way it's really hitting high temperatures.
BTW, due to a related Xen issue, after S3 the information is reported only for CPU0. After fixing that, it properly gets reported for other CPUs too, and for all of them it ends up in "modulated clock mode".
How reproducible
Always?
How to reproduce
- Start Qubes OS 4.3
- Suspend
- Resume
- Wait a little (like 30s)
- Observe
xl dmesg
There is also more high level test for the performance degradation after S3, but it involves waiting much longer after resume (at lest 30min): QubesOS/qubes-issues#10616 (comment)
Expected behavior
Ideally no messages about temperature above threshold. But if there is any, there should be also matching message about returning to normal mode.
Actual behavior
Last message is Running in modulated clock mode
Screenshots
No response
Additional context
I see also somebody reported similar issue on NV41, but I haven't verified it there.
Solutions you've tried
No response