On Linux 6.12+ with a grep kstack /proc/vmstat there is a break down of the kernel stack usage in power-of-two buckets. This histogram was added to the kernel with fleets of "millions of machines" on the mind. For hyperscalers and others with vast deployments of Linux servers, small optimizations to memory use can make a profound impact.
This is beneficial to efforts around the dynamic kernel stack efforts and hyperscalers/CSPs with the resources to pursue relentlessly optimizing the Linux kernel.
The full list of MM updates for the Linux 6.12 kernel can be found via this pull request that has already been merged to Linux Git.