LinHES Forums :: View topic - R5F27 system clock going double speed

Sun Sep 23 11:29:21 2007 -1.252517 seconds
Sun Sep 23 11:46:41 EDT 2007

Sun Sep 23 11:29:22 2007 -0.529105 seconds
Sun Sep 23 11:46:43 EDT 2007

Sun Sep 23 11:29:23 2007 -0.390184 seconds
Sun Sep 23 11:46:45 EDT 2007

Sun Sep 23 11:29:24 2007 -0.421265 seconds
Sun Sep 23 11:46:47 EDT 2007

Sun Sep 23 11:29:25 2007 -0.621174 seconds
Sun Sep 23 11:46:49 EDT 2007

Sun Sep 23 11:29:26 2007 -0.726071 seconds
Sun Sep 23 11:46:51 EDT 2007

================
> > From: Gregory Poudrel <chojin at chojin.info>
> > I recently had some problems with my ASUS P5A motherboard (K6-II 350) on
> > FreeBSD 5.2-RELEASE.
> > In fact, I had GENERIC kernel installed and ACPI activated.
> >
> > First problem:
> > Time flow was incorrect. In fact, 1 second meaned 0.5 second for FreeBSD.
> > Each minute I tested a ntpdate and there was 76 seconds difference.
> > When I rebooted and check in BIOS, time was correct and flowed independently.
>
> This is a known bug in the P5A. It's ACPI-Safe clock is broken and runs
> at double speed.

If that is the only problem with it, then it is easy to fix by telling
it what its actual speed is (sysctl machdep.acpi_timer_freq=whatever).
This can be done automatically by calibrating the frequency relative to
any non-broken clock with a known frequency, as is done for the TSC
timecounter.

> Use the TSC clock by putting the line:
> kern.timecounter.hardware=TSC
> into /etc/sysctl.conf.

But using the TSC is better anyway, since it has a higher quality than the
acpi timer if thermal control mechanisms are not used (usually the same
accuracy, and always a lower overhead).
==========================================================================
[503] Thu 30.Dec.2004 16:51:10 [kadmin at archangel][~]
#sysctl kern.timecounter.choice
kern.timecounter.choice: TSC(800) ACPI-fast(1000) i8254(0) dummy(-1000000)

Last box I had trouble like this on was K6-2/500 on Asus p5a mobo.
Setting to i8254 worked in that case. If you've not tried all the
possibilities, do so now.

Author:	guidow [ Sun Sep 23, 2007 12:51 pm ]
Post subject:	R5F27 system clock going double speed
System : ASUS P5A w/ AMD-K6(tm) 3D processor 450 MHz w/PVR-350 using TV-out. No sound card. Promise Tx2 Ultra 133 w/ Seagate 250G drive. I have been using Myth on this set from R5B7 to R5D1 with many hours of trouble free enjoyment even with the underpowered CPU. I did a fresh install of R5F27 and configured for SchedulesDirect. The Problem: The system time is going twice as fast as the hw clock: Running hwclock;date gave me this: Code: Sun Sep 23 11:29:21 2007 -1.252517 seconds Sun Sep 23 11:46:41 EDT 2007 Sun Sep 23 11:29:22 2007 -0.529105 seconds Sun Sep 23 11:46:43 EDT 2007 Sun Sep 23 11:29:23 2007 -0.390184 seconds Sun Sep 23 11:46:45 EDT 2007 Sun Sep 23 11:29:24 2007 -0.421265 seconds Sun Sep 23 11:46:47 EDT 2007 Sun Sep 23 11:29:25 2007 -0.621174 seconds Sun Sep 23 11:46:49 EDT 2007 Sun Sep 23 11:29:26 2007 -0.726071 seconds Sun Sep 23 11:46:51 EDT 2007 What I have done using information from searching the forums: NTP could not seem to keep up and compensate for the rapid change. I have turned off ntpd and confirmed it was not the cultprit. I know enough to check the HW clock and make sure it is set to UTC time. I also deleted /etc/adjtime before reboot to make sure the time reported off the HW clock was not skewed. I also added the following to lilo.conf, even though I am dealing with the nforce problem: Code: append = "nomce apm=power-off quiet noapic noacpi noioapic" I also tried to add pci=noacpi got a boot warning that with argument was not valid. Further googles indicate a problem with the MB that have never shown up before: Quote: ================ > > From: Gregory Poudrel <chojin at chojin.info> > > I recently had some problems with my ASUS P5A motherboard (K6-II 350) on > > FreeBSD 5.2-RELEASE. > > In fact, I had GENERIC kernel installed and ACPI activated. > > > > First problem: > > Time flow was incorrect. In fact, 1 second meaned 0.5 second for FreeBSD. > > Each minute I tested a ntpdate and there was 76 seconds difference. > > When I rebooted and check in BIOS, time was correct and flowed independently. > > This is a known bug in the P5A. It's ACPI-Safe clock is broken and runs > at double speed. If that is the only problem with it, then it is easy to fix by telling it what its actual speed is (sysctl machdep.acpi_timer_freq=whatever). This can be done automatically by calibrating the frequency relative to any non-broken clock with a known frequency, as is done for the TSC timecounter. > Use the TSC clock by putting the line: > kern.timecounter.hardware=TSC > into /etc/sysctl.conf. But using the TSC is better anyway, since it has a higher quality than the acpi timer if thermal control mechanisms are not used (usually the same accuracy, and always a lower overhead). ========================================================================== [503] Thu 30.Dec.2004 16:51:10 [kadmin at archangel][~] #sysctl kern.timecounter.choice kern.timecounter.choice: TSC(800) ACPI-fast(1000) i8254(0) dummy(-1000000) Last box I had trouble like this on was K6-2/500 on Asus p5a mobo. Setting to i8254 worked in that case. If you've not tried all the possibilities, do so now. I can't find any site that can tell what these timer options mean and if they might offer a solution.

Author:	guidow [ Mon Sep 24, 2007 7:15 pm ]
Post subject:
Looks like this is the magic combination that has fixed the problem with the acpi=off being the active ingredient: Code: append = "nomce apm=power-off quiet acpi=off"

LinHES Forums http://forums.linhes.org/

R5F27 system clock going double speed http://forums.linhes.org/viewtopic.php?f=2&t=16693	Page 1 of 1

Page 1 of 1	All times are UTC - 6 hours
Powered by phpBB® Forum Software © phpBB Group http://www.phpbb.com/