Clock drift on Transmeta Crusoe

Clock drift on Transmeta Crusoe

[Michèl Alexandre Salim]

Hello,

Searching through the mailing list I could not find a
reference to this problem, hence this post.

Having ran various kernel and distribution
combinations (SGI's 2.4.2-xfs bundled with their Red
Hat installer, 2.4-xfs-1.0 and 2.4 CVS trees, Linux
Mandrake with default kernel 2.4.3, and lastly
2.4.5-ac9), compiled for generic i386 and/or Transmeta
Crusoe with APM off or on, one thing sticks out : a
clock drift of a few minutes per day.

This problem might not be noticeable for most users
since notebooks are not normally left running that
long, but it is rather serious. I can choose not to
sync the software and hardware clock on shutdown and
re-read the hardware clock every hour or so but it is
rather kludgy.

Any suggestions and/or user experiences more than
welcome.

Regards,

Michel

____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie

Re: Clock drift on Transmeta Crusoe

[Pavel Machek]

HI!

> Searching through the mailing list I could not find a
> reference to this problem, hence this post.
> 
> Having ran various kernel and distribution
> combinations (SGI's 2.4.2-xfs bundled with their Red
> Hat installer, 2.4-xfs-1.0 and 2.4 CVS trees, Linux
> Mandrake with default kernel 2.4.3, and lastly
> 2.4.5-ac9), compiled for generic i386 and/or Transmeta
> Crusoe with APM off or on, one thing sticks out : a
> clock drift of a few minutes per day.
> 
> This problem might not be noticeable for most users
> since notebooks are not normally left running that
> long, but it is rather serious. I can choose not to
> sync the software and hardware clock on shutdown and
> re-read the hardware clock every hour or so but it is
> rather kludgy.
> 
> Any suggestions and/or user experiences more than
> welcome.

Let me guess: vesafb?

If problem goes away when you stop using framebuffer (i.e. go X), then
it is known. 

You are lucky. My machine is able to loose 2 minutes from every 3
minutes.

try time cat /etc/termcap, and check it against stopwatch.
 								Pavel
-- 
I'm pavel@ucw.cz. "In my country we have almost anarchy and I don't care."
Panos Katsaloulis describing me w.r.t. patents at discuss@linmodems.org

Re: Clock drift on Transmeta Crusoe

[Jonathan Morton]

>> clock drift of a few minutes per day.

That's about 0.1%.  It may be relatively large compared to tolerances of
hardware clocks, but it's realistically tiny.  It certainly compares
favourably with mkLinux on my PowerBook 5300, which usually drifts by
several hours per day regardless of actual load.

The drift might be caused by something masking interrupts for too long, too
often, considering you state that the hardware clock remains comparatively
well-synced.  As another poster suggests, the framebuffer may be to blame.

--------------------------------------------------------------
from:     Jonathan "Chromatix" Morton
mail:     chromi@cyberspace.org  (not for attachments)

The key to knowledge is not to rely on people to teach you it.

GCS$/E/S dpu(!) s:- a20 C+++ UL++ P L+++ E W+ N- o? K? w--- O-- M++$ V? PS
PE- Y+ PGP++ t- 5- X- R !tv b++ DI+++ D G e+ h+ r++ y+(*)

Re: Clock drift on Transmeta Crusoe

[Alan Cox]

> > Mandrake with default kernel 2.4.3, and lastly
> > 2.4.5-ac9), compiled for generic i386 and/or Transmeta
> > Crusoe with APM off or on, one thing sticks out : a
> > clock drift of a few minutes per day.
> Let me guess: vesafb?
> 
> If problem goes away when you stop using framebuffer (i.e. go X), then
> it is known. 

2.4.5-ac has the console irq disaster fixed so I'd expect it to seem ok. 


Alan

Re: Clock drift on Transmeta Crusoe

[Michèl Alexandre Salim]

--- Jonathan Morton <chromi@cyberspace.org> wrote: >
>> clock drift of a few minutes per day.
> 
> That's about 0.1%.  It may be relatively large
> compared to tolerances of
> hardware clocks, but it's realistically tiny.  It
> certainly compares
> favourably with mkLinux on my PowerBook 5300, which
> usually drifts by
> several hours per day regardless of actual load.
Several hours a day, gosh...

Thanks for the responses, is it a common problem in
notebooks then? Did not notice this on desktops
before, anyway trying to adjust for the drift using
adjtimex now.

Regards,

Michel

____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie

Re: Clock drift on Transmeta Crusoe

[Michèl Alexandre Salim]

--- Pavel Machek <pavel@suse.cz> wrote:
> Let me guess: vesafb?
I am running vesafb, yes...

> If problem goes away when you stop using framebuffer
> (i.e. go X), then
> it is known. 
but the problem happens in X as well :)

> You are lucky. My machine is able to loose 2 minutes
> from every 3
> minutes.

Indeed :) Thanks, it seems like mine is just a normal
drift.

Regards,

Michel

____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie

Re: Clock drift on Transmeta Crusoe

[Pavel Machek]

Hi!

> > Let me guess: vesafb?
> I am running vesafb, yes...
> 
> > If problem goes away when you stop using framebuffer
> > (i.e. go X), then
> > it is known. 
> but the problem happens in X as well :)

So that's different problem.

> > You are lucky. My machine is able to loose 2 minutes
> > from every 3
> > minutes.
> 
> Indeed :) Thanks, it seems like mine is just a normal
> drift.

My 2-minutes-from-3-lost problem is caused by heavy scrolling in
vesafb. It is known console bug. X can not cause that!
								Pavel
-- 
I'm pavel@ucw.cz. "In my country we have almost anarchy and I don't care."
Panos Katsaloulis describing me w.r.t. patents at discuss@linmodems.org