High CPU Temperature
After updating the office desktop to Slackware 15.0, an oddball quirk surfaced with sensors. Sometime the Conky display showed an absurd high CPU temperature. Checking the underlying sensors command confirmed the unbelievable value.
The motherboard is an ASUS Z170-K with a 2.7 GHz quad core i5-6400 CPU.
The displayed temperature would jump from about 30 to about 110 degrees Celsius. At that high of a temperature the system fans should be rattling and the BIOS should shut down the system.
Reloading the sensors command always provided remedy. More odd is the bug seems to occur only once per boot.
Searching the web found many people asking about high CPUTIN temperatures. Many people replied that the sensor is unreliable and should be ignored because this is a motherboard sensor rather than a built-in CPU sensor. This Linux kernel document explains that:
On various ASUS boards with NCT6776F, it appears that CPUTIN is not really connected to anything and floats, or that it is connected to some non-standard temperature measurement device. As a result, the temperature reported on CPUTIN will not reflect a usable value. It often reports unreasonably high temperatures, and in some cases the reported temperature declines if the actual temperature increases (similar to the raw PECI temperature value - see PECI specification for details). CPUTIN should therefore be be ignored on ASUS boards. The CPU temperature on ASUS boards is reported from PECI 0.
That might be so but does not explain why the random spikes never occurred with Slackware 14.2 and the 4.4 kernel.
If this particular sensor is unreliable then what is the actual sensor input to monitor CPU temperature? Newer Intel CPUs use the Platform Environment Control Interface (PECI). Running the sensors command shows some related CPU core temperatures. None explicitly are identified as PECI 0, but one of the values is Package id 0. This value seems to be a weighted average. This is the only related CPU value that matches the BIOS display. Updating the conkyrc file to use that value stopped the bogus warnings.
Newer is not always better.
Yet Another Example why updating computers is so frustrating.
Something always breaks. Always.
Posted: Category: Usability Tagged: Slackware
Next: BUG: unable to handle page fault for address
Previous: Intel CPU Frequency Governors