i have 8 cores and for some reason i get only the first core getting overheat… is there anything i get do about it so it will spread the load across other cores too? i mean so not only the first core will get overheating all the time…
How are you checking your cores temperature?
Do you have some hardware monitoring IC on your motherboard that is showing you the CPU cores temperature? Because you can’t read the cores temperature directly from CPU in dom0 because of Xen restrictions.
1 Like
i use sensors and it actually does show me the cores and their temperatures…
Maybe you’ve pinned some specific core to dom0?
Can you show the output of sensors to see what temperatures do you have?
1 Like
sorry but i don’t want to show my specs if this is ok… but i didn’t do anything to attach specific core to dom0… i didn’t even know it’s possible to do that…
How much does “overheating” mean? 5°C? 10°C? 40°C? And what’s the average temperature for other cores?
1 Like
normally i have 40-45c but when overheating only the first core is getting to 75 and above… the rest is also getting hotter but not that much… only till 50…
Check the output of this command in dom0:
sudo xl dmesg | grep "Command line"
Do you have dom0_max_vcpus or dom0_vcpus_pin options there?
Also is it only one specific core that is overheating or the core that is overheating can change?
1 Like
thank you for trying to figure out this with me and help but no… i don’t have this commands… there’s some unknown things there like gnttab_max_… ucode… etc and memory… but at least now we know i don’t have some core attached to dom0…
Is it only one specific core that is overheating or the core that is overheating can change?
If you have some process in one of your qubes that is 100% of the core then it could lead to this specific core overheating.
I don’t know how Xen CPU scheduler works but it could be that the core assigned to this running at 100% process is not changing until the process is finished and this will lead to this core overheating.
1 Like
that’s the thing that it’s only the first one… the temp1 that is getting hot.
i don’t have anything that is going even near the 100% on any qube… never… and it doesn’t seem like it’s about some specific qube either… and i don’t know really what to look for to troubleshoot this properly… i feel like this is not healthy for the cpu but honestly not really sure either
Are you sure that it’s a core temperature and not some other sensor temperature like GPU temperature or something?
I don’t know where are you checking the temps and how are they named.
1 Like
yeah pretty sure… i have fans speeds there too… and other data… and it shows temp like temp1 and its temp temp2 and its temp etc…
I’d recommend reading the documentation on the “sensors” command to learn what information it specifically displays. For example, temp1 would be a specific thermistor reading and not CPU core reading. The manual for the motherboard will go into more detail on the thermistors installed and purpose(s). The BIOS would show which cores are active (default is all).
ahh… OK thank you… i will try to not rely on these but i do get my laptop hotter when these numbers get bigger…
For my laptop the temp1 is CPU temperature (sensor near CPU on the motherboard).
Progress!
Then temp1 is for the CPU package and not any individual core (“sensors” would display these as core0, core1, etc., but wouldn’t display them if controlled by ACPI like on many laptops).
Chip designers (ask me how I know) typically target CPU package temperatures to 100C or below to meet reliability requirements. If your system is below this then it is fine. If it is at or above 100C then there is an insufficient cooling [mechanical] issue. This would not be Qubes or any other operating system related. I’d recommend not wasting time with random, not data-driven, speculation. Check out Linus Tech Tips, Tom’s Hardware, Republic of Gamers, etc. sites on how to address overheating issues.