Subject: Re: HEADS UP: timecounters (branch simonb-timecounters) merged into -current
To: None <current-users@netbsd.org>
From: Wolfgang S. Rupprecht <wolfgang+gnus200606@dailyplanet.dontspam.wsrcc.com>
List: current-users
Date: 06/13/2006 16:18:26
Frank Kardel <kardel@netbsd.org> writes:
> The FreeBSD implementation of Poul-Henning Kamp's timecounters
> was ported and committed to NetBSD-current.

It looks like the backwards compatibility might have a minor quirk.
I'm running a new kernel from this afternoon sand a older userland
from May 24.  The processor is a vintage Intel PPRO-150.

    NetBSD cayenne.wsrcc.com 3.99.21 NetBSD 3.99.21 (WSRCC_PPRO) #0: Tue
    Jun 13 15:33:15 PDT 2006
    wolfgang@capsicum.wsrcc.com:/var/obj/netbsd/sys/arch/i386/compile/WSRCC_PPRO
    i386

    $ sysctl kern.timecounter
    kern.timecounter.choice = clockinterrupt(q=0, f=100 Hz) TSC(q=800, f=149704260 Hz) i8254(q=100, f=1193182 Hz) dummy(q=-1000000, f=1000000 Hz)
    kern.timecounter.hardware = TSC
    kern.timecounter.timestepwarnings = 0

    $ ntpq -p
         remote           refid      st t when poll reach   delay   offset  jitter
    ==============================================================================
    xcapsicum.wsrcc. .GPS.            1 u   19   64  377    0.484  -252.60  86.778
    xbonnet.wsrcc.co 25.96.172.67     2 u   28   64  377    0.455  -391.58 106.604
     LOCAL(0)        LOCAL(0)        12 l   22   64  377    0.000    0.000   0.015

Notice the 'x' in front of the hostname signifying a lost lock.  It
was locked to the first host for a while but the offset climbed from
the single digits and the lock was eventually lost.  Is this expected?

In any case, its probably not that important.  At some point the
userland will finish building and things will probably work much
better then.

-wolfgang
-- 
Wolfgang S. Rupprecht                http://www.wsrcc.com/wolfgang/