Lost time?

Lost time?

[- Reyneke]

Hi everyone=2C

After a move from the ppc (v2.6.19) branch to the powerpc (v2.6.25) we've s=
potted a bit of an annoying problem. It looks like we loosing system time o=
n a running system. About 0.6sec per minute.

It's not hardware=2C Old kernel was rock solid with regards to time. Even w=
ith a stripped down kernel (minimum device drivers=2C modules=2C and IRQ's)=
 we still see time loss. During kernel debug all tick calculations looks re=
asonable.

Anyone have any ideas=2C obvious suspects=2C or pointers where to go and lo=
ok?

Hardware is a 500Mhz ppc 440EPx based system.

 Regards

_________________________________________________________________
100=92s of Nikon cameras to be won with Live Search
http://clk.atdmt.com/UKM/go/101719808/direct/01/=

Re: Lost time?

[Scott Wood]

- Reyneke wrote:
> Anyone have any ideas, obvious suspects, or pointers where to go and look?

Check the kernel's idea of what the timebase frequency is on both kernels.

-Scott

Re: Lost time?

[- Reyneke]

> Check the kernel's idea of what the timebase frequency is on both kernels=
.

> check for changes in the U-Boot and your u-Boot environment variables=2C =
device tree bolb content=2C etc=2C not just the kernel


Thanks for all the feedback. This going to hurt=2C but the problem was emba=
rrassingly simple. Our CONFIG_SYS_CLK_FREQ in u-boot was wrongly defined (w=
e did a u-boot upgrade as well - 1.3.3). Instead of 33000000 (doubled 16.5M=
hz input clock) it was defined as 33333333. And there is your 0.6sec for ev=
ery 60sec...

Using 20/20 hind site this is the first thing I SHOULD have checked=2C but =
there you go.

 Regards
   Jan Reyneke

_________________________________________________________________
The John Lewis Clearance - save up to 50% with FREE delivery
http://clk.atdmt.com/UKM/go/101719806/direct/01/=