Re: [patch 2/2] x86, fpu: lazy allocation of FPU area - v3

[Andi Kleen]

On Fri, Mar 07, 2008 at 07:20:44AM -0600, Arjan van de Ven wrote:
> Andi Kleen wrote:
> >>this is just for handling the case where that fails
> >>(basically near/totally OOM or the case where you get a fatal signal)
> >
> >I didn't think GFP_KERNEL was interruptible by signals...
> >(although sometimes under oom thrashing I think it would be great if it 
> >was...) 
> 
> we need to make it (or with GFP_KILLABLE); would make total sense...
> (so yeah it was more wishful thinking than reality)

I think it wouldn't be that difficult for the normal anonymous user allocations
(standard page fault path), but doing it for everything would be pretty
hard because you would need to add signal-bail-out paths everywhere. 

But doing it for some simple cases like page fault only would be a nice 
project for someone, shouldn't be too difficult.

-Andi
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/