You are not logged in.
Hi,
I've noticed that my CPU fan was almost always on recently, using kernel 5.4 (5.4.3 to be precise). Especially as compared to 4.19-LTS.
I reproduced exact same use condition ( = my daily work).
I finally did
- reboot, read temperatures using 'sensors' and 'tlp-stat'
- reboot to LTS, do the same
- etc, 3 times
Here are the temp readings 1 minute after reboot :
5.4 :
coretemp-isa-0000
Adapter: ISA adapter
Package id 0: +77.0°C (high = +100.0°C, crit = +100.0°C)
Core 0: +77.0°C (high = +100.0°C, crit = +100.0°C)
Core 1: +73.0°C (high = +100.0°C, crit = +100.0°C)
BAT0-acpi-0
Adapter: ACPI interface
in0: 8.30 V
curr1: N/A
acpitz-acpi-0
Adapter: ACPI interface
temp1: +25.0°C (crit = +107.0°C)
pch_skylake-virtual-0
Adapter: Virtual device
temp1: +67.5°C
CPU temp = 89 [°C]
4.19 :
coretemp-isa-0000
Adapter: ISA adapter
Package id 0: +54.0°C (high = +100.0°C, crit = +100.0°C)
Core 0: +54.0°C (high = +100.0°C, crit = +100.0°C)
Core 1: +56.0°C (high = +100.0°C, crit = +100.0°C)
acpitz-acpi-0
Adapter: ACPI interface
temp1: +25.0°C (crit = +107.0°C)
pch_skylake-virtual-0
Adapter: Virtual device
temp1: +51.0°C
CPU temp = 63 [°C]
any idea what to do / maybe where to file a bug ?
Using Corei7 7th Gen on XPS 13.
Last edited by squalou (2019-12-30 17:26:23)
Offline
Hi,
I can confirm exactly the same thing with exactly the same machine.
It seems to be related to Gnome Shell because the problem disappears when using Xfce.
Offline
Hi,
I must disagree : I'm using Xfce myself.
Thanks for the feedback anyway, this is becoming very, very weird. Interestingly weird but weird
With X, not wayland, in case that matters. (with Xfce I'm not even sure there's another choice)
I wonder what else could make any difference.
SSD is crypted, and I saw that "aes_x86_64" module has disappeared from kernel 5.4, supposedly by something more efficient.
Aside from that .... no idea, really.
Offline
You are right, I didnt test enough but I can confirm the same problem with xfce.
It's a really annoying problem.
When I downgrade to 5.2 or 5.3 the problem still occurs whereas I never had the problem when I was using those kernels two or three months ago.
Offline
so, would mean it's a compbination of kernel > some-version + something else ...
damn
If anyone knowns how tmepreatures sensors work, will be welcome !
Offline
Have you tried to use i7z and see what temp it gives you?
Offline
Didn't know the tool.
Here's teh output with 4.19 kernel
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % C7 % Temp VCore
Core 1 [0]: 1188.78 (11.89x) 12.8 88.8 1.89 1.82 1.16 54 0.7067
Core 2 [1]: 1162.86 (11.63x) 5.11 11.4 1 17.4 67.7 52 0.7064
be right back
Last edited by squalou (2019-12-21 18:01:12)
Offline
with 5.4.5
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % C7 % Temp VCore
Core 1 [0]: 1161.43 (11.62x) 3.18 8.55 1 13.4 75.5 51 0.7020
Core 2 [1]: 1125.07 (11.26x) 13.8 78.3 1.07 3.21 10.9 50 0.7017
rebooting back again to check
Offline
I did several attemps, in same conditions (with / without power cable plugged in (which causes hiher temp) ), a few minutes after reboot or immediately.
Seems consistent.
Finally after many mor reboots :
with 4.19 and 5.4.5 both
- sensors, and i7z, all say around 50 degrees
- tlp-stat say 70 degrees
maybe was fixed in last kernel patch ? (initally I had issues wuth 5.4.3)
Compared to my initial post, looks like 'sensors' output is fixed now.
fan issue .... to be confirmed
Last edited by squalou (2019-12-21 18:10:37)
Offline
It looks like your readings are much more consistent with i7z than using lm_sensors.
sensors can have errors and you can adjust the output:
https://wiki.archlinux.org/index.php/Lm … and_tricks
Offline
I' m far too lazy to rollback to kernel 5.4.3.
But apparently, problem is solved.
I've quickly checked the Changelog on kernel.org, one commit referred to something related to temperature. Maybe it was that.
Anyway, so far, so good, let's close this thread !
Offline
Hmm,
back at work, problems are back.
Here are the readings : with 5.4.5 and 10 minute safter boot, pc idle.
Core [core-id] :Actual Freq (Mult.) C0% Halt(C1)% C3 % C6 % C7 % Temp VCore
Core 1 [0]: 1531.33 (15.31x) 12.3 15 1.09 13.6 62.4 70 0.7830
Core 2 [1]: 1512.53 (15.13x) 12.6 16.5 1 8.36 66.2 68 0.783
ONE difference is : I'm using a usb-c dock (a monitor acting as a dock, to be precise)
In this situation, temperature never goes lower than 67 degrees.
Whereas with 4.19, it goes own to 50 degrees
.... so, tha would mean that the real issue is with the usb-c with charging and display etc etc magic, that somehow causes issues ?
I also notice that when using kernek 5.4.5 with this dock, the battery charge indicator never reaches 100% (it stays at 98%)
as if something was draining battery perpetually... but it's not the CPU. Maybe the USB-C subsystem somehow ?
I must admit I'm lost, probably not a CPU reading issue, maybe a real temperature issue but not from where I thought it was ....
I remove the 'SOLVED' tag for a time, in case someone steps by,
but I'll probably close / reopen a new thread once I know best what happens.
Offline