https://bugzilla.novell.com/show_bug.cgi?id=350981
User robin.listas@telefonica.net added comment
https://bugzilla.novell.com/show_bug.cgi?id=350981#c4
Carlos Robinson changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEEDINFO |NEW
Info Provider|robin.listas@telefonica.net |
--- Comment #4 from Carlos Robinson 2008-01-03 18:26:42 MST ---
Ok, first I report my findings. I booted my spare 10.2 partition
by the way: I need more than 15 partitions, and today doings
proves the point: I have 10.2, 10.3, 10.3 Beta, windows, etc.
;-)
I will add the logs later. First an excerpt:
minas-tirith:~ # grep -i "timer\|clock\|acpi_pm\|tsc" /var/log/boot.msg
<6>ACPI: PM-Timer IO Port: 0x4008
<4>Calibrating delay using timer specific routine.. 3603.71 BogoMIPS
(lpj=7207432)
<6>..TIMER: vector=0x31 apic1=0 pin1=2 apic2=-1 pin2=-1
<7>PCI: Setting latency timer of device 0000:00:1e.0 to 64
<6>Machine check exception polling timer started.
<6>Real Time Clock Driver v1.12ac
<6>Time: tsc clocksource has been installed.
<6>Time: acpi_pm clocksource has been installed.
<3>i8xx TCO timer: failed to reset NO_REBOOT flag, reboot disabled by hardware
<7>PCI: Setting latency timer of device 0000:00:1f.5 to 64
<6>intel8x0_measure_ac97_clock: measured 59323 usecs
<6>intel8x0: clocking to 48000
<notice>boot.clock start
Setting up the hardware clockdone
<notice>'boot.clock start' exits with status 0
minas-tirith:~ #
(no complain of tsc unstable)
I left the machine running for about 20', the ntp daemon didn't complain. Then
I stopped it and played around.
First, I changed the clock to 'pit':
minas-tirith:~ # cat
/sys/devices/system/clocksource/clocksource0/current_clocksource
acpi_pm
minas-tirith:~ # echo pit >
/sys/devices/system/clocksource/clocksource0/current_clocksource
minas-tirith:~ # cat
/sys/devices/system/clocksource/clocksource0/current_clocksource
pit
minas-tirith:~ #
minas-tirith:~ # date && sleep 5 && date
Fri Jan 4 01:19:37 CET 2008
Fri Jan 4 01:19:42 CET 2008
You see that the clock is running fine with 'pit'. The same test in 10.3 locks
completely suse 10.3 requiring hard reset.
Then I try 'jiffies':
minas-tirith:~ # echo jiffies >
/sys/devices/system/clocksource/clocksource0/current_clocksource
minas-tirith:~ # cat
/sys/devices/system/clocksource/clocksource0/current_clocksource
jiffies
minas-tirith:~ #
minas-tirith:~ # date && sleep 5 && date
Fri Jan 4 01:21:04 CET 2008
Fri Jan 4 01:21:09 CET 2008
The same change in my 10.3 partition stops the clock completely. The system
doesn't crash right away, but tasks needing a time delay do crash (like "top").
I think this already proves there is a software problem: it is the same
machine, simply another partition booted with 10.2. Both 'pit' and 'jiffies'
fail in 10.3 and work fine in 10.2
I also tried 'tsc' and 'acpi_pm':
minas-tirith:~ # echo tsc >
/sys/devices/system/clocksource/clocksource0/current_clocksource
minas-tirith:~ # cat
/sys/devices/system/clocksource/clocksource0/current_clocksource
tsc
minas-tirith:~ # date && sleep 5 && date
Fri Jan 4 01:22:11 CET 2008
Fri Jan 4 01:22:16 CET 2008
minas-tirith:~ # echo acpi_pm >
/sys/devices/system/clocksource/clocksource0/current_clocksource
minas-tirith:~ # cat
/sys/devices/system/clocksource/clocksource0/current_clocksource
acpi_pm
You see that all clock sources work fine in 10.2
Now, I try a 10 minutes wait, with 'acpi_pm' as the clocksource:
minas-tirith:~ # echo ; ntpdate -q pool.ntp.org && sleep 600 && echo && ntpdate
-q pool.ntp.org
server 195.10.6.126, stratum 2, offset 0.025518, delay 0.13368
server 158.227.99.106, stratum 1, offset 0.020131, delay 0.10342
server 80.38.21.80, stratum 3, offset 0.023261, delay 0.12755
4 Jan 01:25:34 ntpdate[4606]: adjust time server 158.227.99.106 offset
0.020131 sec
server 195.10.6.126, stratum 2, offset 0.018281, delay 0.13515
server 158.227.99.106, stratum 1, offset 0.013015, delay 0.10312
server 80.38.21.80, stratum 3, offset 0.015666, delay 0.12764
4 Jan 01:35:35 ntpdate[4653]: adjust time server 158.227.99.106 offset
0.013015 sec
minas-tirith:~ #
That clocksource in my 10.3 has visible delays.
/var/log/messages:
Jan 4 01:18:45 minas-tirith kernel: Time: pit clocksource has been installed.
Jan 4 01:18:56 minas-tirith /usr/sbin/gpm[4251]: *** info [startup.c(145)]:
Jan 4 01:18:56 minas-tirith /usr/sbin/gpm[4251]: Started gpm successfully.
Entered daemon mode.
Jan 4 01:20:59 minas-tirith kernel: Time: jiffies clocksource has been
installed.
Jan 4 01:22:03 minas-tirith kernel: Time: tsc clocksource has been installed.
Jan 4 01:24:11 minas-tirith zmd: Daemon (WARN): Not starting remote web server
Jan 4 01:24:38 minas-tirith kernel: Time: acpi_pm clocksource has been
installed.
Another thing: the problem reported in Bug 350980 (which might be related to
this one) also disappears in 10.2
I will add now the 10.2 partition boot logs
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.