The attached patch is sufficient to make coreinfo run on a machine without TSC, however while trying to get FILO to run I've realised this involves more changes than I originally expected.
cpu_khz gets exposed as a global, so it's going to be very difficult to support CPUs without TSC, without changing the mechanisms some payloads use to do timing. If you don't have TSC, then cpu_khz is largely irrelevant, but FILO accesses it directly.
I think this global needs to be removed and some other mechanism introduced for handling timeouts.
regards, Biff.