Intel has upstreamed some Resource Control “resctrl” improvements to Linux 7.0 for enhanced telemetry monitoring. This is the good kind of telemetry with this new code being useful for being able to monitor how much energy or work is attributed to a group of tasks / process IDs on the system.
The x86/cache pull request for the Linux 7.0 merge window comes down to the Intel resctrl additions for being able to more easily monitor energy use for a defined group of tasks:
“Extend the resctrl machinery to support telemetry monitoring on Intel. The practical usage of this is being able to tell how much energy or how much work can be attributed to a group of tasks tracked under a single idenitifier. Prepend this work with proper refactoring of resctrl domains handling code. Work by Tony Luck.”
Tony Luck at Intel has been working on this resctrl telemetry monitoring for quite a while. After 17 rounds of patch review, it’s finally merged for Linux 7.0. This telemetry monitoring does depend upon hardware capabilities found within Intel CPUs.
The patch cover letter provides more background on the telemetry monitoring capabilities as well as some examples ofr setting up resctrl for monitoring based on a list of process IDs. From there via resctrl sysfs is energy monitoring targeting that group of tasks and the like.
It’s a useful addition for Linux 7.0 for those wanting to have more detailed energy reporting based on processes with modern Intel platforms.
