For those using the generic ACPI Collaborative Processor Performance Control (CPPC) CPUFreq driver such as on a number of ARM64 Linux server platforms, if the CPUs support a "boost" mode it's disabled by default. But a patch from Huawei is now looking to enable CPU boost mode by default on Linux.
The boost mode for processors where the highest performance level is greater than the nominal performance level has been disabled by default with the ACPI CPPC driver and required setting the "boost" option via sysfs. Especially for Linux server platforms, it could be considered a poor default where peak performance potential is typically desired.
This patch proposes enabling boost by default to avoid the need for user-space to set it manually.
Arm engineer and Linux developer Christian Loehle responded to the patch with:
"I often wonder if we should keep boost altogether, the power and thermal envelope on silicon nowadays is complex enough for this to have little meaning, it in no way guarantees no throttling or that you receive the requested performance. The (silicon-aware) user can always individually set the ceilings s.t. it 'works' for their platform/workload/environment combination via scaling_max_freq.Since commit 54e74df5d76d was also introduced by Huawei I'd be curious about the usage there, AFAICS having a platform with NOMINAL_PERF != HIGHEST_PERF is just additional headache on linux for now?"
So we'll see what happens with the patch or ultimately how the Linux ACPI CPPC code is cleaned up for helping with modern ARM64 servers.