I had to reset
/sys/devices/system/clocksource/clocksource0/current_clocksource (in
this case to jiffies).
To list available sources:
cat /sys/devices/system/clocksource/clocksource0/available_clocksource
Do you know of a link or file with documentation on each type of clock?
Sorry, found a link on a xen mailing list (server is a Xen Dom0 host),
but haven't kept it.
I'll try brute force...
nimrodel:~ # echo tsc >
/sys/devices/system/clocksource/clocksource0/current_clocksource
nimrodel:~ # cat
/sys/devices/system/clocksource/clocksource0/current_clocksource
tsc
I have set clocksource=jiffies as a boot parameter. After this all my
ntp time sync problems disappeared.
I'm not sure... I just updated the kernel tonight, and 45' after boot
I see:
Dec 6 22:14:08 nimrodel kernel: Clocksource tsc unstable (delta =
65620380907 ns)
Dec 6 22:51:06 nimrodel kernel: set_rtc_mmss: can't update from 0 to 51
Dec 6 22:51:09 nimrodel syslog-ng[3946]: last message repeated 2 times
Dec 6 22:51:09 nimrodel kernel: set_rtc_mmss: can't update from 1 to 51
Dec 6 22:51:35 nimrodel syslog-ng[3946]: last message repeated 22 times
Dec 6 22:51:36 nimrodel kernel: set_rtc_mmss: can't update from 1 to 51
Lots of those messages till 23:09:54, then they stopped. Maybe the
'tsc' clock is bad and it switches to 'acpi_pm'. Perhaps I'll have to
try them all.
Can't comment on tsc, I've never used it. However, on a laptop with a
new 10.3 install (standard kernel) acpi_pm did give me the same problems
- changing to jiffies solved the problem for me.
________________________________________________________________________
This e-mail has been scanned for all viruses by Star. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
http://www.star.net.uk
________________________________________________________________________
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]