...  There's nothing stopping you from just updating the i386-specific
CLKF_INTR to use your new global variable, and committing this (or
sending a change-request PR) for the i386. that would be Really Cool.

I dont see how this'd extend to a multiprocessor; is a global variable
really the right thing for CLKF_INTR()?  In general, I dont think so.

And somene should fix systat ps, so that instead of the braindead
load-average "histogram" it shows interrupt and system time. :)