NSLU2-Linux
view · edit · print · history

Info.TheNSLU2TimeOfDayTimekeeping History

Hide minor edits - Show changes to markup

August 13, 2007, at 02:23 AM by fcarolo -- fixed false wikilinks
Changed lines 1-2 from:

The NSLU2 has a hardware Real Time Clock (RTC), however while Linux is running, time is actually kept using interrupts from the built in XScale? clock registers (i.e. the RTC is not used).

to:

The NSLU2 has a hardware Real Time Clock (RTC), however while Linux is running, time is actually kept using interrupts from the built in XScale clock registers (i.e. the RTC is not used).

Changed lines 23-24 from:
20.107OpenDebianSlug? Brian Wood, OpenDebianSlug?
9.715UnSlung? 5.5 Beta5 daysChris Burghart, chris_burghart
to:
20.107OpenDebianSlug Brian Wood, OpenDebianSlug
9.715Unslung 5.5 Beta5 daysChris Burghart, chris_burghart
August 11, 2007, at 06:06 PM by Thomann -- Added drift entry
Changed lines 30-31 from:
-0.528Debian 2.6.18.dfsg.1-10 / Etch rc120 daysmas
to:
-0.528Debian 2.6.18.dfsg.1-10 / Etch rc120 daysmas
11.187OpenSlug 3.10216 daysThomann
February 22, 2007, at 09:20 PM by mas --
Changed line 30 from:
-0.528Debian 2.6.18.dfsg.1-10 / Etch rc120 days|mas
to:
-0.528Debian 2.6.18.dfsg.1-10 / Etch rc120 daysmas
February 22, 2007, at 09:19 PM by mas -- added mine
Changed lines 29-30 from:
10.269OpenSlug 3.101 day 7hBernard Davison
to:
10.269OpenSlug 3.101 day 7hBernard Davison
-0.528Debian 2.6.18.dfsg.1-10 / Etch rc120 days|mas
February 22, 2007, at 07:12 AM by Bernard Davison -- Added ntp.drift info
Changed lines 28-29 from:
7.49OpenSlug 2.7189 daysTim Auton
to:
7.49OpenSlug 2.7189 daysTim Auton
10.269OpenSlug 3.101 day 7hBernard Davison
August 06, 2006, at 01:57 PM by Tim Auton -- Added NTP drift results
Changed lines 27-28 from:
10.727OpenSlug 2.7 Adrian Day
to:
10.727OpenSlug 2.7 Adrian Day
7.49OpenSlug 2.7189 daysTim Auton
February 17, 2006, at 05:11 PM by Peter Taylor -- inserted a comma after running
Changed lines 1-2 from:

The NSLU2 has a hardware Real Time Clock (RTC), however while Linux is running time is actually kept using interrupts from the built in XScale? clock registers (i.e. the RTC is not used).

to:

The NSLU2 has a hardware Real Time Clock (RTC), however while Linux is running, time is actually kept using interrupts from the built in XScale? clock registers (i.e. the RTC is not used).

January 09, 2006, at 01:12 PM by jbowler -- Add/update data
Changed line 22 from:
217.246OpenSlug 2.7 Brian Wood, OpenSlug
to:
-26.371OpenSlug 2.735Brian Wood, OpenSlug
Changed lines 26-27 from:
15.123OpenSlug 2.726 dayspTweety
to:
15.123OpenSlug 2.726 dayspTweety
10.727OpenSlug 2.7 Adrian Day
January 09, 2006, at 10:08 AM by beww --
Changed line 22 from:
217.246OpenSlug Brian Wood, OpenSlug 2.7
to:
217.246OpenSlug 2.7 Brian Wood, OpenSlug
January 09, 2006, at 10:07 AM by beww --
Changed line 22 from:
217.246OpenSlug Brian Wood, OpenSlug
to:
217.246OpenSlug Brian Wood, OpenSlug 2.7
January 08, 2006, at 03:09 PM by jbowler -- added data
Added line 19:
6.818OpenSlug 3.02 daysJohn Bowler, lochalsh
January 06, 2006, at 01:19 PM by pTweety --
Changed lines 24-25 from:
48.680OpenSlug 2.5113 daysRoger Nilsson
to:
48.680OpenSlug 2.5113 daysRoger Nilsson
15.123OpenSlug 2.726 dayspTweety
January 06, 2006, at 08:40 AM by Chris Burghart -- Updated table to include my name and new ntp.drift (with more time to stabilize)
Changed line 23 from:
10.292UnSlung? 5.5 chris_burghart
to:
9.715UnSlung? 5.5 Beta5 daysChris Burghart, chris_burghart
January 06, 2006, at 04:49 AM by Birko Bergt --
Changed line 19 from:
20.145OpenSlug Birko Bergt, slugracer
to:
20.145OpenSlug 2.740 daysBirko Bergt, slugracer
January 05, 2006, at 11:37 AM by nilrog -- Added my ntp.drift info
Changed lines 23-24 from:
10.292UnSlung? 5.5 chris_burghart
to:
10.292UnSlung? 5.5 chris_burghart
48.680OpenSlug 2.5113 daysRoger Nilsson
January 04, 2006, at 11:12 AM by jbowler -- fix formatting
Changed lines 17-23 from:
Drift |! OS Version |! uptime |! name
9.872|OpenSlug 2.7|4 days|John Bowler, cuillin
20.145|OpenSlugBirko Bergt, slugracer
7.293|OpenSlug 2.7Jochen Rüter
217.246|OpenSlugBrian Wood, OpenSlug
20.107|OpenDebianSlug?Brian Wood, OpenDebianSlug?
10.292|UnSlung? 5.5chris_burghart
to:
DriftOS Versionuptimename
9.872OpenSlug 2.74 daysJohn Bowler, cuillin
20.145OpenSlug Birko Bergt, slugracer
7.293OpenSlug 2.7 Jochen Rüter
217.246OpenSlug Brian Wood, OpenSlug
20.107OpenDebianSlug? Brian Wood, OpenDebianSlug?
10.292UnSlung? 5.5 chris_burghart
January 04, 2006, at 11:10 AM by jbowler -- Added some data
Added lines 15-23:
NSLU2 clock drift samples
Drift |! OS Version |! uptime |! name
9.872|OpenSlug 2.7|4 days|John Bowler, cuillin
20.145|OpenSlugBirko Bergt, slugracer
7.293|OpenSlug 2.7Jochen Rüter
217.246|OpenSlugBrian Wood, OpenSlug
20.107|OpenDebianSlug?Brian Wood, OpenDebianSlug?
10.292|UnSlung? 5.5chris_burghart
January 04, 2006, at 10:53 AM by jbowler -- Brief description of how the XScale HW clock works
Added lines 1-14:

The NSLU2 has a hardware Real Time Clock (RTC), however while Linux is running time is actually kept using interrupts from the built in XScale? clock registers (i.e. the RTC is not used).

The mainstream versions of the Linux kernel assume that IXP4XX boards provide a clock input to the CPU which causes clock counts to happen at a rate of 66.666666MHz. Unfortunately this would require a 33.333333MHz crystal and no one seems to sell those...

In practice, therefore, an IXP4XX board has some built in clock drift which can be determined from the actual clock frequency. On the NSLU2 board the crystal is 33MHz and the clock count rate is 66MHz, other IXP4XX boards use 33.33MHz or 33.3333MHz crystals which are much closer to the Linux expected value. The Intel specified value is actually 33.33MHz (though the documentation may not be consistent.)

OpenSlug kernels correct for the actual clock frequency in a variety of ways (depending on the kernel). In addition NTP software (ntp or openntp) will automatically correct the clock from external time server sources.

NTP (but not openntp) writes a file /etc/ntp.drift which is the clock error in parts per million (PPM). Over time (at least 24 hours) this value will give the actual accuracy of the crystal on the specific board. By collecting values from different boards we can work out how reproducible the crystals are.

In is my [jbowler's] idea that all NSLU2 boards show approximately the same drift and that some internal inaccuracy in the OpenSlug software is causing a small (about 10ppm) drift. If this can be established as a consistent error we can improve clock accuracy on the NSLU2 in the absence of NTP.

Therefore if you install NTP on openslug (or variants) and run it for long enough (at least 24 hours) to both produce /etc/ntp.drift and get a consistent value please enter the value below:

view · edit · print · history · Last edited by fcarolo.
Based on work by Thomann, mas, Bernard Davison, Tim Auton, Peter Taylor, jbowler, beww, pTweety, Chris Burghart, Birko Bergt, and nilrog.
Originally by jbowler.
Page last modified on August 13, 2007, at 02:23 AM