Comment 2 for bug 364656

Does getconf CLK_TCK also report 94?

There is a relevant debian bug for this; http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=460331 which at least explains why this is reported as invalid.

A common way to compute this is to sleep 1 second and count the difference in timer interrupt events. procps similarly uses /proc/stat to get timer jiffy events, and /proc/uptime to get total system uptime. Roughly, by dividing jiffy events by uptime gives procps the tics per second value it uses, rather than any kernel HZ value actually set. In any case, however it is being derived, any HZ value between 66-94 ticks/second is currently considered invalid in procps, and any value between 95-105 is silently adjusted to 100. This 94 report could be some result of clock skewing under QEMU. The 2.6.30 arm kernel I use on qemu/versatile reports 100.