Re: page allocation failures in mv643xx_eth_poll (was Re: Swapper problems?)
On Sun, Nov 22, 2009 at 01:24:01PM +0000, David Given wrote:
> > Did anyone figure this one out? I'm getting these regularly on
> > my Sheevaplug.
> Unfortunately, no; I've been asking about this for years, both on the
> NSLU2 and the SheevaPlug, but never really got much in the way of a
> I *think* it's a bug that's causing it to spuriously trigger an
> out-of-memory condition under high load situations --- possibly it's
> running out of free pages but is in some state that's preventing it from
> discarding used pages to make more?
When the code that's asking for more memory is running from atomic
context, the system can't sleep to free up more memory.
I.e., an atomic allocation is "Either give me the memory now or tell
me you don't have any, but don't tell me to wait."
There's no bug anywhere, this is just expected behavior.