This is AFAIK standard, documented 2.4.0 kswapd behavior. Again AFAIK. Check
the main kernel mailing lists, I believe it was discussed there.
On Sunday 28 January 2001 12:20 am, John G Dorsey wrote:
> [EMAIL PROTECTED] wrote:
> > The CPU % of kswapd is almost 100% ! That means kswapd is a
> > compute-intensive program?
>
> One could probably argue about the amount of useful computation being
> done by kswapd in this case. =)
>
> The `top' output really suggests that kswapd has run amok on your
> system. Unfortunately, I'm afraid I can't think of an occasion when I've
> seen that symptom, so perhaps someone else on the list will be able to
> suggest a diagnostic. (I could imagine kswapd doing a large amount of
> work in situations where the system has become -- in error -- nearly
> starved for allocatable pages. For this situation to occur immediately
> upon startup seems unlikely, unless some logic error, say, in the
> initial conditions for the zoned allocator, has been introduced into
> your kernel.)
>
> For reference, it may be helpful to know if this condition is repeatable
> with a less dated kernel.
>
> -jd
>
> _______________________________________________
> http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm
> Please visit the above address for information on this list.
--
Edward Muller - [EMAIL PROTECTED] - [EMAIL PROTECTED]
Grow a ponytail -- view it as your telepathic antenna to other Linux Kernel
Developers. -- Jeff V. Merkey
_______________________________________________
http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm
Please visit the above address for information on this list.
