On Sat, Jan 15, 2000 at 06:28:40PM +0100, Kevin D. Kissell wrote:
> As Ralf says, it's very accurate *if* you have an accurate picture of
> the CPU's clock frequency. The usual mechanism of measuring
> the progress of the count register against elapsed time on a TOD
> clock has been known to be inaccurate on some platforms due
> to skew between the visibility to software and the actual timing
> event, so be careful.
The only machine where we actually calibrate the CPU counter is the Indy
where due to a hardware bug we cannot use the external timers as source
In fact the current Linux kernel would even panic if it ever would find
an Indy working as per spec ;-)
We could improve that meassurement by rounding to the nearest crystal
known to be used on IP22 CPU modules plus maybe add an eventual user /
xntpd suplied fudge factor for deviations of the exact crystal frequency.