Re: 2.5.7[23]: wall-clock time advancing too rapidly?

[john stultz <johnstul@us.ibm.com>]

On Mon, 2003-06-23 at 17:16, Andy Pfiffer wrote:
> Hmmm... I tried the patch in 2.5.73 and it appeared to have no effect.

Ah, just when I sent it off to Andrew. Well, I've been getting a number
of successful reports, so its still good to give it further testing. 


> The system continues to advance what it thinks is wall-clock time by
> about 7.25 seconds for every 15 seconds actual wall-clock time:
[snip]
> 
> None of the printk's in the patch have been printed on the console.

Yea, it clearly isn't triggering the code. 

> Adding "clock=pit" continues to work as a workaround.
> 
> I have attached the output of dmesg from the boot of the kernel with the
> patch present.

Looking over it again you're still not showing any of the signs of
changing cpu-frequency. But the symptoms are very similar. I'm curious,
this is the x220? Do you have a service processor installed in that box?
Maybe we're running into some sort of SMI trouble?

thanks
-john