[ntp:questions] Re: kernel precision on Linux

Piotr Trojanek ptrojane at mion.elka.pw.edu.pl
Fri Jan 30 13:00:10 UTC 2004


On 2004-01-29, Nicholas Garfield <Nicholas.Garfield at cern.ch> wrote:
> Hello,
>
> I am running RedHat 7.2 with all updates.  Following the instructions of FAQ
> Section 6.4 I have established that the kernel is most likely not running a
> nanokernel, nor the pps kit (looking in timex.h).
> This agrees with when I run ntptime and check the status  (0x1 (PLL)) and
> also ntpdc.  kerninfo reports precision at 1microsec, sysinfo reports
> precision at -17 and "system flags:         auth monitor ntp kernel stats".

ntptime and other NTP tools won't report NANO-flag if they were compiled
on system with no NANO-flag in .h files.

> The FAQ then continues to talk about linux differences........ How much can
> these differences affect the precision of a stock redhat box when running
> ntpd v4.

Even running ntpd on linux nanokernel ntpd knows only about microsecond
precision, which comes from gettimeofday() syscall. I'm not sure if it is a
bug, but I patched my linux NTP sources with this patch:

http://home.elka.pw.edu.pl/~ptrojane/systime.diff

and it works quite well (on nanokernel). Now it use nano-precision for all
the timestamps (including NTP requests).

> So I guess I am running the old kernel implmentation with the new daemon
> ntp4.2.  Does anybody know by how much this will affect time precision?

The rule is: use NTP compiled with STA_NANO support on nanokernel and
NTP compiled without STA_NANO on stock kernel:)

-- 
Piotr Trojanek



More information about the questions mailing list