

From there Intel Linux developers also got roped in over the lack of MSR documentation concerning under-volting and any guidance from Intel about any sane manner for supporting such functionality under Linux similar to Intel XTU on Windows. Ultimately though most developers seem in agreement that it would be better for a kernel driver to handle the Intel CPU undervolting than a program running in user-space that is directly interacting with the MSR in question.

The discussion among kernel developers to this patch then evolved into whether touching this undocumented MSR should be permitted and in what scenarios, among other concerns. This patch stems from the recent work on Linux aiming to tighten up access to CPU MSRs from user-space. Jason Donenfeld who is known for his WireGuard work happened to send a patch to kernel developers so writes to the particular OC mailbox MSR won't result in kernel warnings being printed, the particular MSR used by intel-undervolt. That program in turn is touching the CPU MSRs directly for manipulating the behavior. That intel-undervolt functionality relies on reverse-engineering and discoveries made by the community for the support. Besides dropping the CPU voltage, it also allows manipulating the throttling power/thermal limits for Intel processors. There does exist the intel-undervolt program that is unofficial and developed by an independent developer for undervolting Intel CPUs from Haswell and newer on Linux. But a hypothetical Linux kernel driver could be coming for filling such void. While the Intel Extreme Tuning Utility (XTU) on Windows allows for undervolting laptop processors, currently on Linux there isn't any Intel-endorsed way for undervolting your CPU should you be interested in better thermal/power efficiency and other factors.
